Conception of antenna systems for monitoring and

localization applications in the industrial Internet of

things
Amina Benouakta

To cite this version:

Amina Benouakta. Conception of antenna systems for monitoring and localization applications in
the industrial Internet of things. Electromagnetism. Université Côte d’Azur, 2024. English. �NNT :
2024COAZ4011�. �tel-04684048�

HAL Id: tel-04684048

https://theses.hal.science/tel-04684048v1
Submitted on 2 Sep 2024

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
 THÈSE DE DOCTORAT
Conception de Systèmes Antennaires
pour Applications de Supervision et de
Localisation dans l'Internet des Objets
Industriel

Amina BENOUAKTA
Laboratoire d’Électronique, Antennes et Télécommunications (LEAT)

Présentée en vue de l’obtention Devant le jury, composé de :

du grade de docteur en Électronique Thierry Monédière, Professeur des Universités,
d’Université Côte d’Azur Laboratoire XLIM, Université de Limoges.
Laurent Cirio, Professeur des Universités,
Laboratoire ESYCOM, Université Gustave Eiffel.
Dirigée par : Robert Staraj
Philippe Pannier, Professeur des Universités,
Co-dirigée par : Fabien Ferrero Laboratoire IM2NP, Polytech Marseille.
Co-encadrée par : Leonardo Lizzi Robert Staraj, Professeur des Universités,
Soutenue le : 17 Avril 2024 Laboratoire LEAT, Université Côte d’Azur.
Fabien Ferrero, Professeur des Universités,
Laboratoire LEAT, Université Côte d’Azur.
Leonardo Lizzi, Maître de Conférences,
Laboratoire LEAT, Université Côte d’Azur.
 iii

Conception de Systèmes Antennaires pour

Applications de Supervision et de Locali-
sation dans l’Internet des Objets Industriel
Jury :

• Président :

Philippe PANNIER, Professeur des Universités, Laboratoire IM2NP, Polytech Mar-

seille.

• Rapporteurs :

Thierry MONÉDIÈRE, Professeur des Universités, Laboratoire XLIM, Université de

Limoges.
Laurent CIRIO, Professeur des Universités, Laboratoire ESYCOM, Université Gus-
tave Eiffel.

• Examinateurs :

Robert STARAJ, Professeur des Universités, Laboratoire LEAT, Université Côte d’Azur.
Fabien FERRERO, Professeur des Universités, Laboratoire LEAT, Université Côte
d’Azur.
Leonardo LIZZI, Maître de Conférences, Laboratoire LEAT, Université Côte d’Azur.
 v

“Move from reason that observes to reason that acts.”

Haruki Murakami
 vii

UNIVERSITÉ CÔTE D’AZUR

Résumé

Conception de Systèmes Antennaires pour Applications de Supervision et de

Localisation dans l’Internet des Objets Industriel

Cette thèse s’inscrit dans le contexte de l’Internet des Objets (Internet of Things, IoT),
de l’identification et de la traçabilité d’objets dans des environnements dits com-
plexes, réalisées au moyen de la technologie ULB (Ultra-Large Bande) dite de haute
précision temporelle. L’ambition est de contribuer à l’évolution des systèmes ULB
de localisation en temps réel par la conception et l’optimisation d’antennes ULB qui
soient reconfigurables, multistandards et multifonctionnelles. Ainsi, tout système
de localisation intégrant des antennes optimisées sera doté d’une qualité de locali-
sation meilleure et de fonctions nouvelles.
Les contributions principales développées dans cette thèse ont consisté en l’apport
d’améliorations aux systèmes de localisation en temps réel (RTLS) basé sur la tech-
nologie ULB : conception et fabrication d’antennes ULB reconfigurables en fréquence
; conception et fabrication d’une carte électronique de localisation multistandards
(ULB et Long Range - LoRa) ; étude expérimentale des systèmes RTLS intégrant les
antennes conçues et validation de l’évolution de la localisation en termes de portées
supérieures, de détectabilité d’objets sans connaissance préalable de leurs orienta-
tions, et en précision améliorée par l’atténuation de signaux multi-trajets.

Mots clés : Antennes, localisation, systèmes de localisation en temps-réel (RTLS),

ultra-large bande (ULB), lecteur, tag, long range (LoRa), internet des objets.
 UNIVERSITÉ CÔTE D’AZUR

Abstract

Conception of Antenna Systems for Monitoring and Localization Applications in

the Industrial Internet of Things

This thesis is part of the concept of the Internet of Things (IoT), object identifica-
tion, and traceability in so-called complex environments through Ultra-Wide Band
(UWB) technology known for its high temporal precision. The objective is to con-
tribute to the advancement of real-time UWB-based localization systems through the
design and optimization of UWB antennas that are reconfigurable, multi-standards,
and multi-functions. Therefore, any localization system integrating the optimized
antennas will have improved localization quality and new functionalities.
The main contributions developed in this thesis involve enhancements to real-time
localization systems (RTLS) based on UWB technology: design and fabrication of
frequency reconfigurable UWB antennas; design and fabrication of a multi-standard
localization electronic board (UWB and Long Range - LoRa); experimental study of
RTLS systems incorporating the designed antennas and validation of the evolution
of the localization in terms of extended reading ranges, detectability of objects with-
out prior knowledge of their orientations, and improved location accuracy through
the attenuation of multi-path signals.

Key words : Antennas, localization, real-time locating system (RTLS), ultra-wide

bandwidth (UWB), reader, tag, long range (LoRa), internet of things.
 xi

Acknowledgements
First and foremost, I would like to express my sincere gratitude to my supervisors,
Prof. Robert Staraj, Prof. Fabien Ferrero and Prof. Leonardo Lizzi for their support,
valuable guidance and significant contributions to the progression of my scientific
journey during all these past three years. I have learnt not only scientific but also life
lessons, because of their continuous help and patience with me, I will also become a
person who can help others. For this reason, I can never thank them enough.

I would like to express my genuine gratitude to the jury members, Prof. Thierry
Monediere, Prof. Laurent Cirio and Prof. Philippe Pannier, for accepting to review
and evaluate my work and who will be the first people to read these pages. Thank
you.

I would like to thank the LEAT laboratory members and colleagues for all the
meaningful time, and valuable experiences. I wish you all the best in your future
endeavors.

Finally, I want to extend my heartfelt appreciation and gratitude to my family,

my parents and my sister, for their unwavering support during this academic jour-
ney, and without whom I wouldn’t be where I am today. All my achievements in-
cluding these words wouldn’t have become reality without them.
 xiii

Contents

Abstract vii

Acknowledgements xi

Introduction 1

1 Towards high-precision localization of objects 3

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Indoor localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.1 Localization techniques . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.2 Received signal strength-based ranging (RSS) . . . . . . . . . . 3
1.2.3 Time-based ranging . . . . . . . . . . . . . . . . . . . . . . . . . 4
One-way ranging . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Two-way ranging . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Time difference of arrival ranging (TDoA) . . . . . . . . . . . . 5
Angle of arrival localization (AoA) . . . . . . . . . . . . . . . . . 5
1.2.4 Error sources in distance estimation with time-based technique 5
Line-of-sight scenario (LOS) specific error sources . . . . . . . . 6
Non-line-of-sight scenario (NLOS) specific error sources . . . . 6
Error sources in both LOS and NLOS . . . . . . . . . . . . . . . 7
1.3 Indoor localization technologies . . . . . . . . . . . . . . . . . . . . . . . 7
1.3.1 Short range technologies . . . . . . . . . . . . . . . . . . . . . . . 10
1.3.2 Cellular mobile radio localization . . . . . . . . . . . . . . . . . 10
1.3.3 Near Field Technology (NFT) . . . . . . . . . . . . . . . . . . . . 11
1.4 UWB technology for indoor localization . . . . . . . . . . . . . . . . . . 11
1.4.1 History of UWB . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4.2 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4.3 Operating frequency bands of UWB . . . . . . . . . . . . . . . . 12
1.4.4 Principal characteristics of UWB . . . . . . . . . . . . . . . . . . 13
Bandwidth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
The Shannon theorem and UWB . . . . . . . . . . . . . . . . . . 14
Power spectral density . . . . . . . . . . . . . . . . . . . . . . . . 14
High temporal resolution . . . . . . . . . . . . . . . . . . . . . . 15
Robustness against fading cause by multipath propagation . . . 15
Secure and protected communication . . . . . . . . . . . . . . . 15
Obstacle penetration properties . . . . . . . . . . . . . . . . . . . 15
1.4.5 Applications of UWB . . . . . . . . . . . . . . . . . . . . . . . . . 15
Imaging systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Vehicle radar systems . . . . . . . . . . . . . . . . . . . . . . . . 16
Communications systems . . . . . . . . . . . . . . . . . . . . . . 16
Ranging, localization and positioning. . . . . . . . . . . . . . . . 16
1.4.6 UWB Regulations, standards, and consortiums . . . . . . . . . . 16
Regulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
xiv

Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
UWB Consortiums . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.4.7 Conventional UWB antennas . . . . . . . . . . . . . . . . . . . . 21
1.4.8 An overview of conventional UWB antennas . . . . . . . . . . . 22
Mushiake’s principle . . . . . . . . . . . . . . . . . . . . . . . . . 22
Rumsey’s principle . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.4.9 Requirements for UWB antennas in localization systems . . . . 25
The old requirements . . . . . . . . . . . . . . . . . . . . . . . . . 25
The recent requirements . . . . . . . . . . . . . . . . . . . . . . . 26
1.5 Objectives and developed axes of the thesis . . . . . . . . . . . . . . . . 26

2 Time-domain analysis of time-based ranging with UWB antennas 33

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.2 Time-domain analysis principle . . . . . . . . . . . . . . . . . . . . . . . 33
2.2.1 UWB signal processing chain . . . . . . . . . . . . . . . . . . . . 34
2.2.2 The IEEE 802.15.4a complying transmitted UWB signal . . . . . 35
2.2.3 UWB PHY Compliant signal generation . . . . . . . . . . . . . . 36
Nyquist requirement . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.3 Analysis of an antenna operating at one UWB channel . . . . . . . . . . 37
2.3.1 Excitation signal generation and processing using Matlab . . . 38
2.3.2 Antenna characteristics . . . . . . . . . . . . . . . . . . . . . . . 38
2.3.3 EM simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Received E-farfield assessment . . . . . . . . . . . . . . . . . . . 40
2.3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

3 Advancements in Industrial RTLSs: A Technical Review of UWB Localiza-

tion Devices Emphasizing Antennas for Enhanced Positioning 49
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.2 UWB Technology in Real-Time Locating Systems . . . . . . . . . . . . . 49
3.3 UWB Localization from an industry point of view . . . . . . . . . . . . 50
3.3.1 Overview of industry available UWB chips and evaluation boards 51
Qorvo DWM1000 chip . . . . . . . . . . . . . . . . . . . . . . . . 51
STMicroelectronics’ MOD1 chip . . . . . . . . . . . . . . . . . . 54
UWB chips interoperable with Apple’s U1 chip for interactions 56
Murata’s Type2BP-EVK board based on NXP SR150 chip . . . . 59
3.3.2 Summary of the antennas present in industrial UWB chips and
evaluation board . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.4 Discussion of the antenna influence on the localization performance . . 62
3.4.1 Influence on the localization range . . . . . . . . . . . . . . . . . 62
3.4.2 Influence on the tag-orientation dependence . . . . . . . . . . . 64
3.4.3 Influence on the accuracy . . . . . . . . . . . . . . . . . . . . . . 64
3.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

4 Localization with UWB antennas: the requirements of RTLSs 69

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.2 Frequency reconfigurable and circularly polarized patch antenna over
dual Ultra-wideband channels . . . . . . . . . . . . . . . . . . . . . . . 69
4.2.1 Antenna design specifications . . . . . . . . . . . . . . . . . . . . 69
4.2.2 Anchor antenna design principle . . . . . . . . . . . . . . . . . . 72
Design of simple UWB patch at Channel 2 . . . . . . . . . . . . 73
 xv

Design of UWB patch at Channel 5 . . . . . . . . . . . . . . . . . 74

Circular polarization . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.2.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
UWB Frequency reconfigurablity . . . . . . . . . . . . . . . . . . 76
Circular polarization . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.2.4 Experimental characterization of the antenna . . . . . . . . . . . 77
UWB Anchor antenna prototype . . . . . . . . . . . . . . . . . . 77
Measured frequency reconfigurability . . . . . . . . . . . . . . . 78
Measured circular polarization . . . . . . . . . . . . . . . . . . . 78
4.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

5 Antenna Contributions to the Improvement of Localization with UWB Real-

Time Locating Systems 81
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.2 State of the art . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.3 Localization with RTLS technology . . . . . . . . . . . . . . . . . . . . . 84
5.3.1 Ranging method . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.3.2 Instrumentation and configuration . . . . . . . . . . . . . . . . . 84
Transmitter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Communication link . . . . . . . . . . . . . . . . . . . . . . . . . 85
Receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.4 UWB Antenna design . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.4.1 Antenna design motivation . . . . . . . . . . . . . . . . . . . . . 86
5.4.2 Antenna design . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.5 Antenna simulation and experimental results . . . . . . . . . . . . . . . 88
5.5.1 UWB impedance matching . . . . . . . . . . . . . . . . . . . . . 88
5.5.2 Directivity and gain . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.5.3 Radiation pattern and circular polarization . . . . . . . . . . . . 89
5.5.4 Comparison with commercial antenna . . . . . . . . . . . . . . . 90
5.6 Measurement campaign and results . . . . . . . . . . . . . . . . . . . . 92
5.6.1 Outdoor scenario for reading-range . . . . . . . . . . . . . . . . 92
5.6.2 Indoor scenario for object orientation independence . . . . . . . 94
5.6.3 Multipath mitigation with circular polarization . . . . . . . . . 96
5.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

6 Multi-Standard UWB-LoRa High-Accuracy Long-Range Localization 101

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
6.2 Related work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
6.3 UWB-LoRa localization method . . . . . . . . . . . . . . . . . . . . . . . 104
6.3.1 LoRa-UWB sensing and ranging approach . . . . . . . . . . . . 104
6.3.2 Design of the LoRa-UWB transceiver . . . . . . . . . . . . . . . 106
6.3.3 Antenna structure design . . . . . . . . . . . . . . . . . . . . . . 108
6.4 Results and characterization . . . . . . . . . . . . . . . . . . . . . . . . . 110
6.4.1 Antenna measurements . . . . . . . . . . . . . . . . . . . . . . . 110
Impedance matching . . . . . . . . . . . . . . . . . . . . . . . . . 110
Efficiency and gain . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Radiation pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
6.5 Discussion and experimental characterization of the board . . . . . . . 115
6.5.1 From the Sensor-Tag to the Reader . . . . . . . . . . . . . . . . . 115
6.5.2 From the Reader to the Gateway and Network . . . . . . . . . . 117
6.5.3 Localization accuracy . . . . . . . . . . . . . . . . . . . . . . . . 118
xvi

6.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

General conclusion 123

123

Perspectives 125

List of Publications 127

 xvii

List of Figures

1.1 Indoor localization techniques: RSS-based ranging and time-based

ranging (one way, two way and TDoA). . . . . . . . . . . . . . . . . . . 4
1.2 Various LOS and NLOS ranging conditions from a transmitting tag
(TX) to different readers (RXn). RX1 in LOS-scenario, RX2 in NLOS-
scenario with existing direct path and RX3 in NLOS-scenario and com-
plete blockage of the direct path. . . . . . . . . . . . . . . . . . . . . . . 6
1.3 Architecture of the locating systems of different technologies: (a) Short-
range technologies (Wi-Fi, Bluetooth, RFID) [26], (b) Mobile and net-
work based cellular localization [15], (c) Near field technology posi-
tioning system [29]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4 IEEE standard-compliant power spectral density of a UWB transmit-
ted signal, illustrated here at the center frequency of UWB channel
2 [55]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.5 Graphical representation of UWB PHY frequency band allocation [55]. 14
1.6 Power spectral density of the transmit power for different wireless
technologies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.7 Available UWB consortiums names and logos. . . . . . . . . . . . . . . 20
1.8 Envisaged application use-cases for UWB technology by the FiRa Con-
sortium. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.9 Illustration of the Nearby Interaction ranging session with UWB-enabled
Apple devices [60]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.10 Illustration of UWB’s use case: conditionally performing functionali-
ties depending on the presence of a user within the limits of specific
ranging areas [60]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.11 Two self-complementary antennas of different shapes (rectangular and
spiral) and log-periodic antenna in 3D and planar forms. . . . . . . . . 23
1.12 Antennas defined by angles: Spiral antenna, logarithmic spiral an-
tenna and conical spiral antenna. . . . . . . . . . . . . . . . . . . . . . . 23
1.13 Spiral antenna with unbalanced excitation and ground plane. . . . . . 23
1.14 Horn antennas (rectangular and circular cross-sections), TSA anten-
nas of different shapes. . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.15 Evolution from classic dipole antenna to biconical dipole and bowtie
antennas, wire monopole antenna with ground plane orthogonal to
the radiating pole [61]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.16 Planar monopole UWB antennas: two-step band-notch monopole an-
tenna, circular monopole antenna, hexagonal monopole antenna, el-
liptical monopole antenna with truncated ground plane. . . . . . . . . 25

2.1 UWB Transceiver chain composition. . . . . . . . . . . . . . . . . . . . . 34

2.2 Example of a compliant pulse: butterworth pulse p(t) (left), root raised
cosine reference pulse r(t) (middle), and magnitude of their cross-
correlation |ϕ(τ )| (right) [14]. . . . . . . . . . . . . . . . . . . . . . . . . 36
xviii

2.3 The steps of the time-domain analysis of UWB signals’ propagation. . 37

2.4 Construction of the IEEE compliant UWB pulse by RF mixing. . . . . . 38
2.5 Studied UWB patch antenna structure. . . . . . . . . . . . . . . . . . . . 38
2.6 Reflection coefficient of the UWB antenna studied. . . . . . . . . . . . . 39
2.7 Radiation patterns of the UWB antenna, (left): 2-dimensional polar
pattern, (right): 3-dimensional pattern. . . . . . . . . . . . . . . . . . . 39
2.8 Receiver probes along the propagation axis of the antenna, for E-field
observation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.9 Received E-field waveform in frequency domain, and its variation
with distance between the antenna and the probes. . . . . . . . . . . . . 41
2.10 Received E-field waveform in time domain, and its variation with dis-
tance between the antenna and the probes. . . . . . . . . . . . . . . . . 41
2.11 Receiver probes along y-axis of the antenna, for E-field observation. . . 42
2.12 Waveforms of the components of the field (Ex , Ey , Ez ) in the time do-
main, received at one probe. . . . . . . . . . . . . . . . . . . . . . . . . 42
2.13 E-field component waveform Ex (t) and its variation with y-axis coor-
dinates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.14 E-field component waveform Ex (t) and its variation with y-axis coor-
dinates, zoomed-in version. . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.15 Receiver probes around the antenna, for E-field observation. . . . . . . 44
2.16 Received E-field magnitude waveform in time domain, and its varia-
tion with probes at center in front and at the extremities of the antenna. 44
2.17 A, B, C tags yield the same times of arrival. . . . . . . . . . . . . . . . . 45

3.1 DWM1000 frontend monopole ceramic pcb UWB antenna [21]. . . . . . 52

3.2 DWM1000 application board keep-out area for UWB antenna integra-
tion [14]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
3.3 DWM1000 antenna radiation pattern charateristic on azimuth plane
[14]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.4 DWM1000 antenna radiation pattern charateristic on elevation plane
(1) [14]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.5 DWM1000 antenna radiation pattern charateristic on elevation plane
(2) [14]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.6 DWM3000 evaluation board as arduino shield [22]. . . . . . . . . . . . 54
3.7 ESP32 all-in-one evaluation board for UWB tracking [23]. . . . . . . . . 54
3.8 ESP32-Pro (with display) all-in-one evaluation board for UWB track-
ing [23]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.9 STMicroelectronics UWB module MOD1 [20]. . . . . . . . . . . . . . . . 55
3.10 B-UWB-MEK1 application board for MOD1 module with external an-
tenna [24]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.11 MEK1 UWB antenna radiation pattern charateristic on azimuth and
elevation planes [25]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.12 Sunway Communications’ UWB module SR150 interface for ranging
and direction finding [19]. . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.13 Sunway UWB module’s application board SW-EVK-2 with UWB an-
tennas [27]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.14 Sunway Communications’ UWB pcb dual patch antenna [28] for SW-
EVK-2 evaluation board. . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.15 Murata UWB Trimension™ SR150 based application board Type2BP
EVK [29] with UWB antenna : (a) top view, (b) bottom view, (c) per-
spective view. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
 xix

3.16 Murata UWB NXP based modules : (a) Type2BP (built upon the SR150
module), (b) Type2DK (built upon the SR40 module) . . . . . . . . . . . 60
3.17 Range measurement set-up from Murata Type2BP evaluation board
[29]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.18 Accuracy measurement from Murata Type2BP evaluation board [29]. . 64

4.1 8-States frequency reconfigurable UWB monopole antenna [6], with

band rejections at the C band, WLAN, and X band by two C-slots and
inverted U-slot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.2 UWB frequency reconfigurable antipodal Vivaldi antenna [8]: a) op-
eration on UWB mode, b) operation on seven switchable sub-bands. . 71
4.3 Reflection coefficient of the UWB frequency reconfigurable antipodal
Vivaldi antenna: a) operation on UWB mode and 3 sub-bands, b) op-
eration on 7 switchable sub-bands [8]. . . . . . . . . . . . . . . . . . . . 71
4.4 Side view of the UWB anchor antenna. . . . . . . . . . . . . . . . . . . . 72
4.5 Top view of the UWB anchor antenna. . . . . . . . . . . . . . . . . . . . 72
4.6 Bottom view of the UWB anchor antenna. . . . . . . . . . . . . . . . . . 73
4.7 UWB antenna with capacitive feed and air gap. . . . . . . . . . . . . . . 73
4.8 Air gap height effect on bandwidth of Channel 2 . . . . . . . . . . . . . 74
4.9 Channel 2 configuration - Channel 5 configuration. . . . . . . . . . . . 74
4.10 Air gap height effect on bandwidth of Channel 5. . . . . . . . . . . . . . 74
4.11 Feed patch length effect on impedance matching for Channel 5. . . . . 75
4.12 L-shaped slot width effect on impedance matching for Channel 2. . . . 75
4.13 Reflection coefficients of the reconfigurable antenna. . . . . . . . . . . . 76
4.14 Simulated axial ratio of the reconfigurable antenna (main radiation
direction). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.15 Top view of the UWB reconfigurable antenna prototype. . . . . . . . . 77
4.16 Bottom view of the UWB reconfigurable antenna prototype. . . . . . . 77
4.17 Reflection coefficient at channel 2 and channel 5. . . . . . . . . . . . . . 78
4.18 Measured and simulated axial ratio of channel 2. . . . . . . . . . . . . . 79
4.19 Measured realized gain of RHCP and LHCP (at 4 GHz and φ 90°) . . . 79

5.1 Conventional real-time locating system configuration. . . . . . . . . . . 82

5.2 Two-way ranging principle. . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.3 Side view of the frequency reconfigurable UWB antenna [19]. . . . . . 87
5.4 Simplified circular polarized antenna version, operating on one UWB
channel, with radiating element (top layer) and directional coupler
(bottom layer). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.5 Prototype of the circularly polarized antenna operating over UWB
channel 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
5.6 Antenna reflection coefficient. . . . . . . . . . . . . . . . . . . . . . . . . 89
5.7 Measured antenna directivity and gain over the UWB channel 2 fre-
quencies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.8 Measured azimuth antenna radiation pattern (polarisation gain) at
center frequency 4 GHz, for elevation angles 0° and 90°. . . . . . . . . . 90
5.9 Designed circularly polarized antenna’s measured radiation pattern
at frequency 4 GHz in the two polarizations. (a) Elevation 0° (b) Ele-
vation 90°. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.10 Commercial linearly polarized antenna’s measured radiation pattern
at channel 1, 2 and 3 [30]. . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
xx

5.11 Comparison of the RTLS reading range when using the commercial
antenna and the designed antenna. . . . . . . . . . . . . . . . . . . . . . 93
5.12 Two-way ranging with the commercial antenna, characterized by 360°
omnidirectional radiation and linear polarization. . . . . . . . . . . . . 94
5.13 Two-way ranging with the designed antenna, characterized by 180°
directional radiation and circular polarization. . . . . . . . . . . . . . . 94
5.14 Object orientation-independence of received power, comparison be-
tween linearly and circulary polarized antennas. Reader antenna al-
ways in vertical orientation. . . . . . . . . . . . . . . . . . . . . . . . . . 95
5.15 Multipath and antenna gain effects on received signal strength in rang-
ing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

6.1 Communication chain structure from the sensor tags to the network
and gateways. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.2 Organigram of the working localization system highlighting the role
of each node. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.3 Structure of the UWB-LoRa transceiver, composed of two layers, an
electronics bottom layer and an antenna upper layer, connected through
signal vias. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
6.4 Antenna structure design consisting of a UWB patch antenna and a
LoRa PIFA antenna: (a1) top of the upper layer of the board; (b1) bot-
tom of the upper layer of the board; (a2) top of the bottom layer of the
board; (b2) bottom of the bottom layer of the board. . . . . . . . . . . . 108
6.5 Antenna structure design, consisting of a UWB patch antenna and a
LoRa PIFA antenna. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.6 Prototype of the UWB-LoRa transceiver board: (a) the upper layer of
the board containing the ra-diating elements; (b) bottom layer of the
board containing the MCU, the modules and the branch-line coupler. . 110
6.7 Board assembly and measurements: (a) prototyping using foam to
separate the layers and UFL connectors at the coupler inputs; (b) stand-
alone antenna matching measurement set up, using a SMA connector
for the LoRa antenna and UFL-to-SMA cables from the coupler inputs
for UWB antenna characterization. . . . . . . . . . . . . . . . . . . . . . 111
6.8 Reflection coefficient of the LoRa PIFA antenna. . . . . . . . . . . . . . 111
6.9 Reflection coefficient of the UWB patch antenna for both ports of the
coupler (S11 and S22). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
6.10 LoRa antenna’s gain and efficiency characteristics. . . . . . . . . . . . . 113
6.11 UWB antenna’s gain and efficiency characteristics. . . . . . . . . . . . . 113
6.12 Radiation pattern (azimuth) of the UWB antenna at the channel’s cen-
ter and edge frequencies, through polarization realized gain. . . . . . . 114
6.13 Polar radiation pattern of the LoRa antenna at a frequency of 868 MHz. 115
6.14 Three-Dimensional radiation pattern of the LoRa antenna at 868 MHz
frequency, with a view on both left and right sides of the antenna
structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
6.15 Ranging with two LoRa-UWB transceivers, the first being a sensor-tag
and the second being a reader. . . . . . . . . . . . . . . . . . . . . . . . . 116
6.16 Ranging measurements of the UWB localization between the sensor-
tag and the reader, in LOS and NLOS scenarios (outdoor). . . . . . . . 117
6.17 Ranging information of the sensor as displayed after its reception via
the LoRa network. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
 xxi

6.18 Comparison of UWB-measured distances with the true distance be-

tween a reader and a tag in an indoor environment. . . . . . . . . . . . 119
 xxiii

List of Tables

1.1 A qualitative comparison of the available positioning technologies. . . 9

1.2 IEEE standard’s UWB PHY frequency band allocations. . . . . . . . . . 13
1.3 Summary of the evolution of the IEEE standards introducing UWB
technology specifications for precise ranging. . . . . . . . . . . . . . . . 18

3.1 Summary of the UWB radio chips available in the industry, their fre-
quencies and localization techniques. . . . . . . . . . . . . . . . . . . . . 51
3.2 Characteristics of the UWB monopole antenna integrated in the Qorvo
DWM1000 transceiver [21]. . . . . . . . . . . . . . . . . . . . . . . . . . 52
3.3 Characteristics of the UWB 3D monopole antenna of the STMicroelec-
tronics B-UWB-MEK1 evaluation board [24]. . . . . . . . . . . . . . . . 56
3.4 Characteristics of the UWB dual patch antenna [28] of the Sunway
evaluation board [27]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.5 Characteristics of Murata Type2BP module (based on NXP SR150). . . 60
3.6 Characteristics of Murata Type2DK module (based on NXP SR040). . . 60
3.7 Summary of the industrial UWB chips, their evaluation boards and
their antenna integration. . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.8 Summary of the UWB antenna characteristics of the industrial UWB
chips and / or their evaluation boards. . . . . . . . . . . . . . . . . . . . 62
3.9 Range measurement results from Murata Type2BP evaluation board
[29]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

5.1 Adopted instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . 85

5.2 European UWB lowband frequency channels . . . . . . . . . . . . . . . 86
5.3 Antenna radiation comparison . . . . . . . . . . . . . . . . . . . . . . . 91
5.4 Antennae characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

6.1 Summary of the physical dimensions of the antennas. . . . . . . . . . . 109

 1

Introduction
Often, in the telecommunications domain, we find that the word Internet-of-Things
is employed everywhere in literature and applied research, this is because this word
designates any wireless communication between two or several objects or “things”.
This wireless communication consists in the ability of each object to identify another
one, which can either be near or far from it in space. Mostly, and depending on the
type of devices, not only objects can identify each other, but they can also localize
one another. The localization concept is also known as tracking or positioning. In
this context, this thesis addresses mostly the indoor type of localization.
Although outdoor localization is widely known and accessible to the public and
any individual due to the popularity of the Global Positioning System (GPS), indoor
localization however, is more of an in-domain subject. Indoor tracking of things
seems to also be more present in applied industrial settings than in research areas,
especially nowadays, as it has a huge impact in many practices (such as: logistics,
intelligent resource management, supply and stock tracking). Moreover, in research
areas, localization and positioning are more pervasive for fundamental localization
techniques (such as: time-of-flight, time-difference-of-arrival and angle-of-arrival)
or optimization of algorithms relying on probabilistic estimations of object positions.
Even though, less popular, localization in research also concerns the investigation of
the front-end electronics, notably the choices and optimizations of the antennas em-
ployed for radio frequency transmission and reception of the positioning signals.
This thesis exploits and combines both, the research, and industry’s available local-
ization means and solutions to advance towards the improvement of the quality of
localization, more precisely, towards a higher accuracy, a longer reading range, and
detectability without prior knowledge of the target’s orientation.

In the first chapter, an overview of the indoor localization techniques and techonolo-
gies is presented, followed by a more specific view of the Ultra-Wide Band (UWB)
technology which allows for high-precision localization of objects. Then, a state of
the art of the conventional UWB antennas is presented.

In the second chapter, a time-domain analysis of the ranging with UWB signals
and antennas is performed through simulations. First, a UWB signal complying
with the most recent specifications of the UWB standard is generated. Then this
signal is applied to an UWB antenna, and the far-electric field at receiver probes is
studied. The time of arrival of the signals at each probe is then extracted and com-
pared between the receivers. This time is is the time used to determine the distance
in localization.

In the third chapter, a technical review of the existing industrial solutions for lo-
calization is realized. These solutions are real-time locating systems (RTLSs) based
on UWB technology. Then, the characteristics of the antennas integrated in these
systems are highlighted. After that, a discussion about the performance results of
these industrial RTLSs is realized. Notably, the discussion is focused on the influ-
ence of the antennas on these results in terms of range and accuracy.

In the fourth chapter, the design and conception of a UWB antenna is presented.
This antenna is frequency reconfigurable and circularly polarized over two UWB
channels. The design specifications of the antenna are derived from the observa-
tions made in the previous chapter (third chapter). These specifications are set to
2

make the antenna suitable for recent UWB requirements on the bandwidth and al-
low a better reading range and accuracy than conventional UWB antennas.

In the fifth chapter, the previously designed UWB antenna is employed in an in-
dustrial RTLS to perform ranging between a reader and a tag. The objective is to
highlight the advantages of this antenna and its role in enhancing the quality of the
localization results compared to the commercial antenna of the RTLS. The influence
of the characteristics of the antenna employed on the results is discussed in detail
and is in terms of reading range, accuracy and multipath mitigation, and detection
of the tag without prior knowledge of its orientation.

In the sixth chapter, a multi-standard localization RTLS transceiver is designed

employing the previously designed antenna. This RTLS works with UWB technol-
ogy and Long-Range (LoRa) techonology to perform both high-accuracy and long
range localization of a sensor. The objective is to overcome the low datarate of LoRa
techonology which is not suited for real-time localization by employing a UWB tech-
nology reader as an intermediate node between a LoRa gateway and a sensor. Rang-
ing is then performed and the system is assessed.

Finally, a general conclusion on the presented work is realized on the main con-
tributions of this thesis, and the foreseen perspectives of the work are highlighted.
 3

Chapter 1

Towards high-precision
localization of objects

1.1 Introduction
This chapter will present an introduction to indoor localization, its techniques, the
error sources that can be encountered in different scenarios (line-of-sight and non-
line-of-sight), and its technologies such as : Bluetooth, Wi-Fi, and Ultra-Wide Band-
width (UWB). Although some of these technologies were not specifically made for
localization purposes, they have already been exploited to locate objects. Further-
more, the characteristics of the UWB technology, its advantages, applications, stan-
dards and consortiums are presented. Moreover, a state-of-the-art of UWB antennas
is realized. Then, the chapter addresses the old and recent expectations of UWB an-
tennas which need to be considered specifically in the field of localization to evolve
towards more optimized and efficient locating systems. Finally, the technological
axes explored in this thesis are briefly highlighted.

1.2 Indoor localization

Localization of objects starts with inferring the distance between two nodes, this
measure is called “ranging” [1, 2]. After that, obtaining the space coordinates of one
of these nodes is possible through methods like triangulation or with combining the
distance measured with an angle-of-arrival measurement.

1.2.1 Localization techniques

First, to perform ranging, there are many techniques relying on the characteristics of
the signals exchanged between a reader and a tag. The most pervasive techniques
are received signal strength-based ranging (RSS-based), or time-based ranging tech-
nique (figure 1.1).

1.2.2 Received signal strength-based ranging (RSS)

This technique analyses the strength of the signal received at the reader side [3–
8]. The strength of the signal is inversely proportional to the distance, that is, the
stronger the signal, the farther it means the object is. The distance estimate is re-
trieved from the signal power strength measure by using an a priori existing prop-
agation model [9], the simplest being the model of propagation of signals in free
space. Another typical one is the two-path model which considers the direct path
and the reflection from the ground path [10]. Indeed, the more paths the model con-
tains, the more accurate the distance estimate will be. However, for the model to
4 Chapter 1. Towards high-precision localization of objects

F IGURE 1.1: Indoor localization techniques: RSS-based ranging and

time-based ranging (one way, two way and TDoA).

contain the most paths possible, knowledge of the specific environment where lo-
calization needs to be performed, is necessary. Thus, this ranging technique is more
complicated to implement if high accuracy of the distance estimates is needed [2].
Furthermore, the strength of the propagating signal is very sensitive to environmen-
tal factors, such as obstacles (partial or complete blockage of the reader and tag’s
line-of-sight (LoS)), multipath reflections and interferences with other existing sig-
nals, as a low received signal power could indicate that the tag is far away from the
reader but it could also indicate that the direct propagation path is blocked by an ob-
stacle [11]. While these factors make RSS-based ranging less reliable for applications
requiring highly accurate position information, this technique is easier to implement
as it typically doesn’t require additional electronics in the chain of the usual radio
propagation systems, moreover, it can operate with different technologies (Wi-Fi [5],
RFID [6], Bluetooth [7]) as long as a wireless connection is implemented.

1.2.3 Time-based ranging

This technique exploits time measurements to infer the distance information be-
tween a reader and a tag [1, 12, 13], this time is referred to as the “time-of-flight”
(ToF) or “time-of-arrival” (ToA) τ = d/c where d is the unknown distance and c
is the speed of light (3.108 m/s). Time-based ranging exists in many forms [14, 15]:
one-way ranging, two-way ranging, time-difference-of-arrival (TDoA), and angle-
of-arrival (AoA), each of which is described below.

One-way ranging
This method requires the internal clocks of the reader and the tag to be perfectly
synchronized [1,2,9]. The tag sends a packet to the reader containing the information
1.2. Indoor localization 5

on time instant t1 , which is the instant at which the tag sent this message. The reader
receives this information at instant t2 , and is thus, able to compute the ToF τ =
t2 − t1 . This method may seem quicker; however, clock differences exist even in
identical electronics modules, which can induce errors in the time measurements
and thus degrade the estimation accuracy.

Two-way ranging
In this method, the constraint of clock synchronization between the reader and tag
is omitted as the time measurement will be mainly performed by the reader. Here
the time measured is the round-trip time (RTT) [1, 2, 9]. For this, the reader trans-
mits a request (or interrogation) packet to the tag, to which, this latter replies by an
acknowledgment packet after a response delay time τd . The RTT is then computed
as RTT = 2τ f + τd , with RTT known as it is the time duration it took from the time
instant the first signal was sent to the time the reply signal arrives to the reader, and
τd known as it mostly is specified by the manufacturer.

Time difference of arrival ranging (TDoA)

This method requires multiple readers which are usually synchronized through a
wired network connection. A tag broadcasts a signal which is received by the readers
at different time instants depending on their position [16–18]. The readers share with
each other these times of arrival and compute the differences between them (TDoA).
To compute the position of the tag with space coordinates, at least three readers
and two TDoA measurements are required. Each TDoA measurement is interpreted
geometrically as a hyperbola formed by a set of points to determine the tag position
coordinates.

Angle of arrival localization (AoA)

This method involves angle measurements and allows for not only ranging (distance
estimation with ToA), but also localization (coordinates estimation) of a target. Con-
versely to the triangulation method which requires at least three readers it is possi-
ble to combine a ToA measurement method with AoA measurements to estimate the
coordinates. The arrival angle can be determined from the TDoA between signals
received by two or more antenna elements of a single reader. For a two radiating
element antenna array spaced by s, the TDoA depends on the AoA as described by
the equation below [19]:
s
τdi f f =sin(ϕ) (1.1)
c
where ϕ is the AoA, τdi f f the TDoA and c the speed of light.

1.2.4 Error sources in distance estimation with time-based technique

The accuracy of the distance or position information is the most important char-
acteristic and metric of locating systems, and more so in indoor localization. This
accuracy depends on the quality of the time measurements realized. Many factors
can degrade this quality and lead to unreliable measurements and thus to their un-
reliable interpretation while computing the distance estimates. Furthermore, the
principal measurement in time-based ranging is the measure of the arrival time of
the direct path (first path) signal between the reader and tag. While the direct path in
6 Chapter 1. Towards high-precision localization of objects

most cases exists, there can be cases where it does exist, but it is in a non-line-of-sight
scenario, that is for example: if the view between the reader and tag is obstructed
(or blocked) by any obstacle (object, walls. . . ). There is also the case where the direct
path signal may not arrive at all, and the first arriving signal will be a reflection from
one of the objects present in the environment. Indeed, the measurement is consid-
ered ideal and most accurate if the first arriving path is a line-of-sight signal, that is,
the signal travels in a straight line through a medium with constant known relative
permittivity such as in air.

F IGURE 1.2: Various LOS and NLOS ranging conditions from a trans-
mitting tag (TX) to different readers (RXn). RX1 in LOS-scenario,
RX2 in NLOS-scenario with existing direct path and RX3 in NLOS-
scenario and complete blockage of the direct path.

The possible error sources in time measurements can be divided to three cate-
gories: error sources present in LOS-scenario, error sources present in NLOS scenar-
ios and error sources present in both LOS and NLOS scenarios, as described in the
following

Line-of-sight scenario (LOS) specific error sources

In this case, no blockage is present in the direct path between the reader and tag.
The first and fastest arriving signal will be the direct path signal and reflections from
walls and objects arrive later (as they travel longer paths). This is the case where the
measurement will be the most accurate.

Non-line-of-sight scenario (NLOS) specific error sources

In the case where there is blockage between the reader and tag, the first arriving path
may be either the direct path signal after traveling through different mediums (non-
constant relative permittivity) or a reflection signal from an object, the faster signal
is the one that will arrive first and this depends on the permittivities encountered
by the signal while traveling, and how far the reflecting object is. In both cases, the
error will be “excess delay”, which is an added positive bias in the time measured
(and thus in the distance estimate). In the case, of the signal being a reflection signal,
the distance traveled is naturally longer and thus produces a higher time of arrival
1.3. Indoor localization technologies 7

value. While, in the case of the signal being the direct path signal obstructed in
its way by an obstacle, it’s the speed that is reduced. Indeed, the speed of electro-
√
magnetic waves traveling in a homogenous material is reduced by ϵr compared
to the speed of light c. The delay ∆τ introduced by a material of thickness dw and
permittivity ϵr is given by [9]:
p dw
∆τ = ( ϵr − 1 ) (1.2)
c
Recent measurement campaigns in indoor environments, resulted in demonstra-
tions that the mean of the ranging error (distance bias) caused by material blockage
is in the order of the thickness of the obstacle.

Error sources in both LOS and NLOS

In all the scenarios, LOS or NLOS, if ranging is performed in cluttered complex en-
vironments, signal reflections will be present, that is: “multipath effect”. If ranging
is performed with narrowband technologies (Bluetooth, Wi-Fi, RFID. . . ), multipath
components (signals arriving via different propagation paths) combine at the reader
side. These signals can usually not be resolved. Their combination results in con-
structive or destructive interference which makes the detection of the direct path
signal difficult. Furthermore, in complete blockage, multipath signals are the only
signals detected, the problem itself is not one reflection (as this only introduces a
positive bias like explained previously) but is that the combination of the signals
changes its original form which makes its recognition at the reader side impossi-
ble [20]. For this reason, UWB technology is privileged in indoor locations because of
its fine time resolution compared to narrowband technologies, which helps resolve
signals from each other to recognize the correct ranging signal. The advantages of
UWB for ranging are further described in detail in the next section.
Other sources of errors include cochannel interference, which is caused by coexisting
wireless systems sharing the same radio band.

1.3 Indoor localization technologies

The indoor environments are very challenging; thus, multiple technologies have
been proposed to cope with their constraints. Indeed, throughout the years, indoor
localization has been studied and performed using several technologies (figure 1.3),
each conforming to different standards and radio frequencies. These technologies
are heterogenous and include notably : Wi-Fi, Bluetooth, RFID, Near Field Commu-
nications (NFC), and UWB.
At the earliest stages of RTLSs, technologies started from exploiting ultra-sound,
sonar (in radar domains), laser from optics (such as telemeters) and vision technolo-
gies (such as in the robotics domain) [1, 14]. However, recently the current trend to
approach tracking and localization problems is to use standard, low cost and already
deployed technologies. Consequently, in the following, we focus only on technolo-
gies based on radio signal exchange.
Just until recently, most wireless communication standards have been designed for
other purposes and are not intended for localization and ranging purposes, however,
this did not prevent their exploitation for positioning at the cost of some localization
performance limitations such as shorter range, low or average accuracy and, small
detectability coverage. Table 1.1 presents a summary of the existing technologies
along with their characteristics allowing a qualitative comparison, the numbers are
8 Chapter 1. Towards high-precision localization of objects

only indicative as such systems are highly sensitive to the environment changes and
also due to the random nature of the propagation of radio signals.

F IGURE 1.3: Architecture of the locating systems of different tech-

nologies: (a) Short-range technologies (Wi-Fi, Bluetooth, RFID) [26],
(b) Mobile and network based cellular localization [15], (c) Near field
technology positioning system [29].
1.3. Indoor localization technologies 9

Technology Measurement Accuracy Advantages Disadvantages

technique
GPS [1, 22] TDoA 10 – 20 m Worldwide No indoor
coverage Ex-
pensive
Galileo [1, 22] TDoA 1–5m Continent No indoor
coverage, Ex-
pensive
A-GNSS TDoA <5m Global or Low indoor ac-
[1, 22] regional curacy
Cellular TDoA / RSSI 50 – 500 m Country Requires syn-
(2G/3G) chronized base
[1, 15], [23]– stations, Low
[26] accuracy
Cellular TDoA / RSSI 20 m Country Requires syn-
(LTE) [1, 15], chronized base
[23]– [26] stations, low
accuracy
Wi-Fi [1, 23, RSS- 1–5m Indoor Low accuracy,
24, 27] fingerprinting usage requires a fin-
low cost, gerprinting
already data base
deployed
in most
infrastruc-
tures
WSN (Zig- RSS / PDoA 1 – 10 m Indoor Low accuracy,
Bee) [1] usage, low short range
cost, low
power con-
sumption
UWB [1]– ToF / TDoA / 0.1 – 1 m Indoor Problems in
[14], [23], [28] AoA usage, high NLOS scenar-
accuracy ios, infrastruc-
ture cost
RFID / Blue- RSS / Proxim- Connectivity Indoor Low accuracy,
tooth [1, 14, ity / PDoA range (< 10 usage Low short range
23] m) cost, low
power con-
sumption,
small size
Near Field Electromagnetic 1–5m Indoor Low frequency,
Technol- Near Field usage, low large antennas,
ogy [29] characteristics cost short distances
LoRaWAN TDoA, RSSI RSSI: 1000 Low power Outdoor usage
[30] – 2000 m consump-
TDoA: 20 – tion, large
200 m coverage

TABLE 1.1: A qualitative comparison of the available positioning

technologies.
10 Chapter 1. Towards high-precision localization of objects

1.3.1 Short range technologies

Standards for technologies that offer short-range connections between objects are
the most widespread, each of them aiming at a different application. Wi-Fi is the
standard for wireless local area network applications (WLAN) [23, 27], Bluetooth
for wireless personal area network applications (WPAN) [14, 23], RFID for radio
frequency identifications [14, 23], and IEEE 802.15.4 / Zigbee for wireless sensor
networks (WSN) [1]. Since these standards were developed for these specific ap-
plications and not for localization purposes, their exploitation for positioning relies
mostly on extracting parameters of the radio signal received [1, 23, 24, 27], such as
signal strength (RSS) or signal phase for example and then on the correlation of
these parameters with the distance between the tag and reader. Consequently, even
though systems employing such standards are capable of tracking tags, high ac-
curacy of the position information is not trivial. Indeed, because of their narrow
frequency bandwidth, multipath effects on the measurements can be very signifi-
cant. The advantage of these technologies is their wide availability and accessibility
almost anywhere indoor, which removes the necessity of implementing new infras-
tructure specifically for the sake of localization. Smartphones are now often utilized
as receivers for both Bluetooth and Wi-Fi signals, following recent advancements.
While, Beacons are small devices for BLE that are mounted on the walls and repre-
sent small base stations.

1.3.2 Cellular mobile radio localization

As cellular systems evolved from a dedicated mobile communication system to an
almost omnipresent system with unlimited coverage anywhere and anytime for any
device, it is no surprise that it additionally finds applications in other domains such
as positioning and localization. All mobile phone tracking methods have in common
that they use the locations of nearby antennas mounted at radio towers to infer the
mobile’s position. The positioning methods with cellular communications is classi-
fied into two categories: Mobile-based and Network-based. In mobile-based posi-
tioning, the target mobile itself computes its location by exploiting signal measure-
ments from terrestrial transmitters and/or satellites. In network-based positioning,
the network calculates the position of the mobile through relative measurements
between the network and target, these measurements can either be performed by
the network or by the target and transmitter to the network. Potentially, 2G/3G
cellular physical (PHY) layer can provide ranging information through signal TOA
estimation (more precisely, observed TDOA), although the relatively narrow band-
width and time signal structure limit the achievable time resolution (1 µs for Global
System for Mobile Communication (GSM), about 200 ns for 3G systems). LTE tech-
nology mostly employs TDoA and offers a tight synchronization between BSs and
the possibility to use wideband signals with low interference, which showed great
improvement compared to previous cellular generations. The deployment of 5G
and the forthcoming 6G communications are expected to enable high precision posi-
tioning due to the adoption of small cells and massive antenna arrays at millimeter
waves. Indeed, 5G operation frequencies are 700 MHz, 3.5 GHz, and millimeter
waves at 26 and 28 GHz. High frequencies allow high accuracy ranging in direct
(more so in LOS scenarios) but will highly suffer from attenuation, multipath and
reflections (especially in NLOS scenarios) [24]. Conversely, lower frequencies are
more robust to attenuation, hence reaching longer coverage, however, multipath ef-
fects can still deteriorate the accuracy of the ranging.
1.4. UWB technology for indoor localization 11

1.3.3 Near Field Technology (NFT)

Near field technology is a low frequency approach for localization, it is recently gain-
ing interest in parallel with the use of arrays of a high number of antennas such as
in applications for 6G. With large arrays, the system operates in the near-field prop-
agation region (Fresnel region) where the wavefront of the electromagnetic field is
spherical. In this context, the curvature of arrival (CoA) is a measure of the spheri-
cal wavefront and can be used to infer the position of a transmitter target [29]. The
CoA depends on the transmitter position and the array geometry. The drawbacks
of NFT for localization is that the state-of-the-art studies usually refer to only very
short distances and that the used antenna array must be large enough to capture the
spherical characteristic of the received wave, which makes the infrastructure not so
practical.

1.4 UWB technology for indoor localization

Over the years, ultrawide bandwidth signals have existed under many different ap-
pellations notably: impulse radio signal, time domain radio signal, baseband radio
signal, non-sinusoidal radio signal; until the US Department of Defense (DoD) first
set the term ultrawide band to this technology in 1989 [31]. An ultrawide band signal
is defined as constituted of a series of short narrow impulses following each other
with very low duty cycles [28, 32].

1.4.1 History of UWB

Although UWB technology officially acquired its name in 1989, the idea of emitting
very short-duration pulses was not new. In fact, we can say radio was born with
UWB; Indeed, its initiation phase began when James Clark Maxwell proposed the
laws of electromagnetism in 1860 by unifying electrodynamics and magnetism into
a single theory. A few years later in 1888, Heinrich Rudolph Hertz, while trying to
verify Maxwell’s equations, demonstrated the existence of electromagnetic waves
and finally, in 1893, he was able to generate short duration pulses using electrostatic
discharges. These were the dominant wave generators for about 20 years after the
first Hertz experiments [28]. Following that, despite of this UWB glimpse, wireless
communications abandoned impulse transmissions and shifted their focus to the use
of sine waves by employing modulated signals on sinusoidal carriers to allow multi-
user coexistence on the shared radio spectrum and, for almost 50 years, pulse radio
was abandoned.
By the beginning of the Cold War, the second phase had in fact begun. Impulse trans-
mission was rediscovered and contributions to the development of UWB signals and
their applications began in the 1960s with the innovation of Harmuth [33–37], Ross
and Robbins [38–40], and Etten [41]; In 1970, implemented systems for radar applica-
tions appeared with the introduction of short-pulse generators using tunnel diodes.
But until the early 1990s, this work remained confidential because it was aimed at
military applications.
Finally, the third phase which consisted in the standardization and commercializa-
tion of UWB, started with the publication of several studies and pioneering works
of Scholtz and Win [42–50] that demonstrated the potential of UWB for broadband
telecommunications. These works design the transmission of information by asso-
ciating several pulses to the same bit according to position and amplitude spread-
ing codes to allow the sharing of the spectrum between several users. Research
12 Chapter 1. Towards high-precision localization of objects

on UWB evolved further with its experimental characterization for communications

and indoor positioning, in [51]. In February 2002, the US FCC (Federal Communica-
tion Commission) adopted the first report and order [52] which authorized the use
and marketing of certain UWB products in the United States subject to precise but
license-free regulations. This step was a real boost to UWB, as a few years later, Eu-
rope and other continents and countries adopted similar regulations, and since then,
a great deal of work has been done in the industrial, academic, and military worlds.

1.4.2 Definitions
A UWB system is defined as a device using a fractional bandwidth of 25 % or
more [53, 54]. The US regulatory authority FCC extends this definition to a broader
category of signals, including signals with a fractional bandwidth (FBW) at -10 dB
of more than 20 % (or with a frequency bandwidth of more than 500 MHz), as illus-
trated in the figure 1.4. The FBW is represented by the equation below [52] :

BW f − fL
FBW = =2 H (1.3)
fc fH + fL
Where f H and f L are, respectively, the upper and lower frequencies of the signal
frequency band at – 10 dB of the maximal power spectral density (DSP) and BW
is the signal bandwidth at this level, f c is the central frequency and Q is classically
defined as the quality factor.

F IGURE 1.4: IEEE standard-compliant power spectral density of a

UWB transmitted signal, illustrated here at the center frequency of
UWB channel 2 [55].

1.4.3 Operating frequency bands of UWB

The RF requirements for the physical layer of UWB technology impose its operation
on specific defined frequency channels, these channels are listed in table 1.2 and
are categorized in three different band groups numbered from 0 to 2 (respectively:
the channel below 1 GHz, the low band channel group and the high band channel
group), each group contains a mandatory channel for UWB systems operating in
one of its frequency channels and the other channels remain optional if multiband
1.4. UWB technology for indoor localization 13

operation is desired. The mandatory channels are channel number 0 below 1 GHz,
the channel number 3 for low band operation and channel number 9 for high band
operation [55].

Band group Channel Center Bandwidth Mandatory /

1,2 number frequency (MHz) Optional
(MHz)
0 0 499.2 499.2 Mandatory be-
low 1 GHz
1 1 3494.4 499.2 Optional
1 2 3993.6 499.2 Optional
1 3 4492.8 499.2 Mandatory in
low band
1 4 3993.6 1331.2 Optional
2 5 6489.6 499.2 Optional
2 6 6988.8 499.2 Optional
2 7 6489.6 1081.6 Optional
2 8 7488.0 499.2 Optional
2 9 7987.2 499.2 Mandatory in
high band
2 10 8486.4 499.2 Optional
2 11 7987.2 1331.2 Optional
2 12 8985.6 499.2 Optional
2 13 9484.8 499.2 Optional
2 14 9984.0 499.2 Optional
2 15 9484.8 1354.97 Optional

TABLE 1.2: IEEE standard’s UWB PHY frequency band allocations.

Furthermore, figure 5.5 represents the graphical representation of the previous

table, highlighting the channel indexes with respect to frequency, where each chan-
nel is shown as a heavy black line centered on the channel’s center frequency. The
length of the lines depicts the channel bandwidth.

1.4.4 Principal characteristics of UWB

With the need to increase the data rate of wireless systems, UWB technology seems
to be an ideal candidate for future radio communication systems for different types
of networks, thanks to its following characteristics:

Bandwidth
The main characteristic of UWB signals is the width of the frequency band they oc-
cupy, and information theory suggests that with the use of an appropriate code, it is
possible to transmit data at a BER (Bit Error Rate) below an arbitrarily low thresh-
old, provided that the data rate at which it is desired to transmit data, is below the
maximum capacity of the transmission channel.
14 Chapter 1. Towards high-precision localization of objects

F IGURE 1.5: Graphical representation of UWB PHY frequency band

allocation [55].

The Shannon theorem and UWB

The advantages and possibilities of UWB can be summarized by examining the fa-
mous Shannon capacity equation. Capacity is important when demanding applica-
tions require higher and higher data rates [28, 56] such as audio and visual applica-
tions. The Shannon equation is expressed as [57] :

S
C = BW log2 (1 + ) (1.4)
N
Where C is the maximum channel capacity [bit/s], BW is the channel bandwidth
[Hz], S is the signal power [W] and N is the noise power [W].
The maximum channel capacity increases linearly with the signal bandwidth BW,
and only logarithmically with the signal to noise ratio NS . From this, it can be de-
duced that UWB systems have great potential for high-capacity wireless communi-
cations.

Power spectral density

The energy used to transmit a wireless signal is non-infinite and, in general, needs to
be as low as possible, especially for today’s consumer electronics. For a fixed sum of
energy one can either transmit a high energy density over a narrow band or a very
small sum of energy densities over a wide band (figure 1.6). For UWB systems, the
energy propagates over a very wide band and, in general, they have very low power
spectral density (0.5 to 1 mW) [28, 53]. UWB radar systems are the major exception
to this general rule. This characteristic is not intrinsic to UWB signals but is imposed
by radio spectrum regulation organizations.
1.4. UWB technology for indoor localization 15

F IGURE 1.6: Power spectral density of the transmit power for differ-
ent wireless technologies.

High temporal resolution

Due to their wide bandwidth, UWB signals have a high temporal resolution, usually
of around one nanosecond. A first implication of this property is related to localiza-
tion: knowing the delay of a signal with an accuracy of about 0.1 to 1 ns [56], it is
possible to obtain information on the position of the transmitter with an accuracy of
3 to 30 cm.

Robustness against fading cause by multipath propagation

In conventional propagation channels, narrowband systems suffer from fading caused
by multipath propagations as these signals often interfere destructively with the de-
sired signal. In the case of pulsed signals, the transmitted waveforms can have a
large bandwidth, so that multipaths having sub-nanosecond delays can be resolved
and added constructively [56].

Secure and protected communication

UWB signals are inherently difficult to detect. This is because they are spread over a
wide band and transmitted at power spectral density levels close to the noise floor
of conventional radio receivers [56]. These characteristics allow the establishment of
secure transmissions with a low probability of detection and interception.

Obstacle penetration properties

UWB signals offer good penetration capabilities through walls and obstacles, espe-
cially at lower frequencies of the spectrum [53]. This allows for good accuracy in
terms of localization and tracking, thus enabling the development of through-the-
wall vision of radar systems [58].

1.4.5 Applications of UWB

As a result of the imposed standard regulations, UWB systems are grouped accord-
ing to their applications into three principal categories [52] :
16 Chapter 1. Towards high-precision localization of objects

Imaging systems
This first category includes imaging with Ground Penetrating Radar Systems (GPRs),
Through-Wall Imaging, In-Wall Imaging, Surveillance Systems and Medical Sys-
tems. The UWB signal can penetrate the ground or a wall and even go through
it to detect what is hidden behind it. It can also be used to measure distances accu-
rately. The same principle applies to the human body. Therefore, the main users of
systems in this category would be specialists in law enforcement, search and rescue,
construction and mining, and geology.

Vehicle radar systems

In this category, UWB devices can detect the location and movement of objects in
the vicinity of a vehicle, enabling features such as anti-collision, improved airbag
activation and suspension systems to better comply with road conditions. With this
technique, many lives can be saved.

Communications systems
An important application in this class is wireless personal area networks (WPANs),
where data is transmitted over distances of 10 m or less. Two types can be dis-
tinguished in this application, High-Data-Rate (HDR) communications, which are
mainly applied in computer networks (wireless USB) and consumer electronic de-
vices (digital TV). The particularity of this class is that it allows for a very high data
transfer rate (ranging from 100 Mbps to 1 Gbps) and a low-cost installation.
The second type is Low-Data-Rate (LDR) communications, which refers to all prac-
tical applications of sensor networks. These sensor networks are used for intelligent
control of lighting and energy in buildings, automation of industrial processes and
storage processes. These applications require the transfer of small amounts of data
between transmitters (1 Mbps). The UWB’s characteristic of low power transmission
allows LDR communications to use network equipment with very small batteries,
minimizing the infrastructure footprint and cost.
Regardless of these data rate considerations, the environments considered are mainly
office buildings, meeting and conference rooms and public places (e.g., airports or
shopping centers).

Ranging, localization and positioning.

Localization is the subject application, which is detailed in this thesis.

1.4.6 UWB Regulations, standards, and consortiums

Regulations
In the same manner for all technologies which are allocated a place in the radio spec-
trum, the regulations imposed depend on the geographical regions and countries in
the world (US, Europe, Canada, Asia. . . ) and on applications. These regulations
specify notably the frequencies and the power spectral masks allowed for transmis-
sion for a certain technology. Thus, when a new radio communication technology
is introduced, spectrum regulation authorities issue strict rules for the emission of
signals, the aim being to protect existing systems from interference.
1.4. UWB technology for indoor localization 17

Although the transmission power of its signals is very low, UWB must also be reg-
ulated because it will occupy frequency bands already allocated to other technolo-
gies. One of the main particularities of UWB is the absence for the need of a license
to access the UWB frequencies, allowing content to be produced and accessed freely.
However, the decisions taken around the emission of UWB signals can vary from
one regulation to another.

• Regulations in the United States (US) : In February 2002, the FCC issued a
first order on UWB technology that allows signals to be transmitted mainly
in the 3.1 - 10.6 GHz band [52]. This part of the spectrum allows the use of a
bandwidth of up to 7.5 GHz. The power spectral density allowed is of - 41.3
dBm/MHz. Although this power level is low, it is compensated by the band
width used, allowing a total power of 0.5 mW to be transmitted.

• Regulations in Europe (EU) : In March 2006, the standards’ authority ETSI

(European Telecommunications Standards Institute) published a decision that
specifies the technical conditions under which UWB devices can operate in
Europe. The decision covers the use of UWB devices in consumer electronics
applications [59]. UWB devices can use the 6 to 8.5 GHz band with a maxi-
mum power spectral density of -4 1.3 dBm/MHz, which is the same level used
in US UWB signals.

• Regulations in Asia : In Japan, the MIC (Ministry of International Affairs and

Communications) proposed the issuance of unlicensed UWB signals in 2006.
Similar to the FCC, regulations allow emissions at - 41.3 dBm/MHz for UWB
devices. Japan has adopted the European model for the 3.4 - 4.8 GHz band.

In conclusion, spectrum masks depend on applications and regions: in Europe and

Asian countries the regulations tend to be stricter, while in the USA and Canada they
tend to be more extensive. Finally, it should be noted that the regulations imposed
by Europe, the United States and Asia have a common band, the 7.25 GHz - 8.5 GHz
band, without any mitigation techniques and which will eventually make systems
completely nomadic from one continent to another.

Standards
To respond to a real demand and to avoid having incompatible UWB systems which
do not use the same transmitted signal waveform, it is important to establish norms
or standards for communication systems using UWB. The IEEE organization han-
dles the development of standards around the UWB technology. Standards mostly
specify frequencies, signal power masks, data rates and application ranges.
UWB was first briefly introduced in the standard IEEE 802.15.03a (standard for high-
rate communications), however it was first associated with high-accuracy localiza-
tion applications relying on impulse radio in the IEEE 802.15.4a standard (standard
for low-rate communications), for ranges of maximum one meter.
In this thesis, we focus on the most recent IEEE standards, notably the two HRP
UWB IEEE 802.15.4a and IEEE 802.15.4z standards, where HRP stands for High-
Rate Pulse (pulse with high repetition frequency). A summary of the evolution of
the IEEE standards introducing UWB technology is presented in table 1.3, and these
standards are discussed in the following.
18 Chapter 1. Towards high-precision localization of objects

IEEE
802.15.4a
• Starting point for UWB standardization.
(2007)
• UWB PHY changed from an Orthogonal Frequency Division Mul-
tiplexing (OFDM)-based data communication to an impulse radio
technology (IR-UWB).

• Focus on low data rate wireless communication and precision

ranging.
IEEE 802.15.4
(2011)
• First amendment to IEEE 802.15.4a (2007) [20].

• Specifies the High-Rate Pulse (HRP) mode of UWB PHY, that is

transmission of pulses at high rate.
IEEE 802.15.4f
(2012)
• Second amendment to IEEE 802.15.4a (2007).

• Specifies the Low-Rate Pulse (LRP) mode of UWB PHY, that is

transmission of pulses at low rate.

• As the maximum transmitted energy remains the same, LRP

mode transmits stronger pulses, but their number is fewer than
the number of pulses transmitted with HRP mode.
IEEE 802.15.4
(2015)
• Specification of two PHY UWB modes: HRP and LRP modes, cor-
responding to the 2011 and 2012 amendments.
802.15.4z
(2020)
• UWB PHY enhancement with two principal contributions: in-
creased integrity and increased accuracy of the ranging measure-
ments.

• Additional coding and preamble options.

• Improvements to existing modulations.

• Addition of the secure fine ranging feature.

TABLE 1.3: Summary of the evolution of the IEEE standards intro-

ducing UWB technology specifications for precise ranging.

First, the IEEE 802.15.4 standard specifies the PHY and MAC layers of Low-Rate
Wireless Personal Area Networks (LR-WPANs). The IEEE 802.15.4 PHY and MAC
layers are used by higher-layer standards, such as ZigBee® and WirelessHart®.
Multiple PHY schemes are specified in different amendments of the IEEE 802.15.4
standard :

• IEEE 802.15.4a introduced a high-rate pulse repetition frequency (HRP) UWB

PHY used for ranging (i.e., localization).

• IEEE 802.15.4f introduced a low-rate pulse repetition frequency (LRP) UWB

PHY used for RFID, ranging, and reduced energy consumption.
1.4. UWB technology for indoor localization 19

• IEEE 802.15.4z introduced new enhanced modes for both the HRP and LRP
UWB IEEE 802.15.4a/f PHYs.

The HRP UWB PHY specifies a channel bandwidth of 0.5 - 1.3 GHz which corre-
spond to pulse durations from 2 ns to as little as 0.78 ns. Since the calculations
used for ranging techniques rely on the time duration of packet transmission, the
extra short pulse duration makes UWB PHYs suitable for ranging applications. A
finer granularity in the time domain translates to smaller errors in distance estima-
tion. UWB specifications which are mostly described in IEEE 802.15.4a standard and
its amendment IEEE 802.15.4z, specify that UWB waveform generation needs to be
standard-compliant with HRP UWB 802.15.4a/z waveforms for three pulse repeti-
tion frequency (PRF) transmission modes (802.15.4a, and 802.15.4z BPRF and HPRF).
For IEEE 802.15.4a, the valid mean PRF values are 3.9, 15.6 or 62.4 MHz. The IEEE
802.15.4z amendment defines these two PRF modes:

• Base pulse repetition frequency (BPRF), where the mean PRF is 62.4 MHz and
the payload data rate is 6.81 Mbps

• Higher pulse repetition frequency (HPRF), where the mean PRF is either 124.8
or 249.6 MHz.

The scrambled timestamp sequence (STS) field is another key feature introduced by
802.15.4z to enhance data integrity. Transmission of the STS field is optional for the
BRPF and HPRF modes.

- Notes, particularities, points in the amendments :

• In the base pulse repetition frequency (BPRF) mode of IEEE 802.15.4a/z, mean
PRF is 62.4 MHz and data rate is 6.81 Mbps.

• The key difference between the BPRF and the HPRF mode is that in BPRF
the PHR and the payload are modulated with the burst position modulation
(BPM) BPSK technique.

• Similarly to the BPRF mode in IEEE 802.15.4z, IEEE 802.15.4a uses the BPM-
BPSK modulation scheme.

• IEEE 802.15.4a has no STS field.

• The IEEE 802.15.4a/z HRP standard specifies a mandatory compliance check

for HRP pulses. Specifically, the cross-correlation between the used pulse and
a root raised cosine pulse with a roll-off factor of 0.5, must be higher than 0.8
for 0.5 ns in the main (central) lobe, and all other side lobes must have cross-
correlation lower than 0.3.

• The IEEE 802.15.4z amendment specifies that transmitted pulses conform a

specific time-domain mask.

• The IEEE 802.15.4 standard specifies a mask for the transmit power spectral
density PSD.
20 Chapter 1. Towards high-precision localization of objects

UWB Consortiums
The official IEEE standardization has opened doors for other standards and consor-
tiums aiming to reunite the different specifications for UWB RTLSs from different
manufacturers and make them as interoperable as possible. The emerging consor-
tiums were mostly aimed at precise ranging, but others are aimed at different appli-
cations. Below, some of these consortiums (figure 1.7) are described:

F IGURE 1.7: Available UWB consortiums names and logos.

• Fine Ranging Consortium (FiRa): This consortium is an industry consortium

which’s main aim is to enable precise location awareness for people and de-
vices. FiRa describes UWB as a technology that is now demonstrating its po-
tential and quickly becoming a vital mainstream wireless technology like Wi-
Fi and Bluetooth, showed by its adoption in mass market consumer products
(figure 1.8) including devices such as smartphones (iPhones, Samsung, Xiaomi,
. . . ) and smart watches (Apple) since 2019. In addition to the IEEE specified
protocol layers for UWB, FiRa has announced that it is preparing a Common
Service Management Layer (CSML) specification layer that will drive the in-
teroperability of FiRa certified devices.

F IGURE 1.8: Envisaged application use-cases for UWB technology by

the FiRa Consortium.

• Car Connectivity Consortium (CCC)®: Conversely to the FiRa consortium that

focuses on leveraging UWB for precise location, the CCC consortium leverages
UWB in the automotive domain and advances technologies for smartphone-
to-car connectivity solutions. The Digital Key 3.0 standard of the CCC, allows
consumers to use their mobile devices, regardless of manufacturer or operating
system type, to access their vehicles easily and securely (handsfree, location-
aware, keyless access). Indeed, it provides location aware features for cars by
enabling them to localize the device related to the specific car and thus, allow-
ing its user the access to it. Its member companies include smartphone and
vehicle manufacturers, automotive tier-1 suppliers, silicon/chip vendors and
many more. CCC and FiRa consortiums collaborate complementarily with the
same goal of UWB interoperability to develop an ecosystem of UWB devices.

• Apple’s Nearby Interactions: Nearby Interaction is an accessory protocol and

industry standard developed by Apple, it can be used through an application
1.4. UWB technology for indoor localization 21

F IGURE 1.9: Illustration of the Nearby Interaction ranging session

with UWB-enabled Apple devices [60].

F IGURE 1.10: Illustration of UWB’s use case: conditionally perform-

ing functionalities depending on the presence of a user within the
limits of specific ranging areas [60].

to acquire the position of devices with a U1 (Apple’s UWB chip) or another

UWB chip, such as iPhone 11 or later, Apple Watch, and UWB enabled third-
party accessories. Apple defines it as lightweight application-level protocol
which specifies the configurations needed to start an interaction session, in
which devices that are in physical proximity can participate (figures 1.9 and
1.10), and share their position and device identity, the application notifies of
the peers of their positions by reporting their direction and distance in meters.

• Omlox Consortium: Omlox name was derived from Latin terms “omni-present”
and “locus” (location), is the first emerging open locating standard dedicated
to RTLSs, it aims to enable each manufacturer to implement its own location-
based services based on the standardized architecture, this architecture intro-
duces the omlox core-zone that ensures interoperability across RTLS and de-
vices from different vendors. Its members include Huawei, Tencent, Haier,
Inspur, and Hikvision.

1.4.7 Conventional UWB antennas

According to the FCC regulation authority and standards defined by the IEEE or-
ganization, a UWB antenna is defined as an antenna having a bandwidth equal or
greater than 500 MHz, i.e., a fractional bandwidth equal or greater than 20 %.
In the first part of this section, an overview of UWB antennas, how they emerged,
and their evolutions are described as generally known; In the second part, details,
22 Chapter 1. Towards high-precision localization of objects

and expectations of UWB antennas are described in the context of localization and
ranging.

1.4.8 An overview of conventional UWB antennas

Similarly with standard narrowband front ends, UWB antennas are designed to op-
erate within the bandwidth of the system where they are planned to be integrated.
Hence, in the past years, UWB antennas were majorly designed to work with UWB
systems which transmit signals over a power mask covering all the 7.5 GHz spec-
trum bandwidth allocated for UWB applications (3.1 to 10.6 GHz). Furthermore,
UWB antennas are no exception to other antennas and need to adhere to the same
design specifications and constraints such as: impedance bandwidth, radiation pat-
tern, directivity and gain, polarization, efficiency, size constraint, cost, etc. However,
the fact that these parameters are frequency dependent make the design procedure
of UWB antennas more difficult and challenging compared to its narrowband coun-
terparts. First, UWB systems transmit signals with either one of these transmission
techniques [61]: pulsed transmission, known as impulse radio UWB (IR-UWB) and
covers all the UWB bandwidth, or multiple broad sub-bands transmission, known
as multiple orthogonal frequency-division multiplexing (OFDM). Catering to each
method, the same constraints apply either in all the ultra-wide frequency bandwidth
or in multiple wide frequency bands, and the principal objectives of UWB antenna
design stand as: steady directional or omni-directional radiation pattern, stable di-
rectivity and gain at the radiation directions of interest, constant polarization type
with respect to frequency, high radiation and total efficiencies, and for IR-UWB sys-
tems, constant group delay of the antenna (to avoid pulse distortions) is added.
Since the emerging pioneering works of Scholtz and Win (Scholtz et al. 2005) on
UWB transmissions, research leaned again towards the design of their antennas. In
fact, UWB antennas belong to the category of the broadband antennas, more com-
monly called: frequency independent antennas.
To design frequency independent antennas, two theoretical principals exist: the
Mushiake principle (1940) and the Rumsey principle (1950) [61].

Mushiake’s principle
Mushiake’s principle is related to the Babinet principle [62] and states that an an-
tenna’s performance will be frequency independent if the antenna is self comple-
mentary. A self-complementary antenna (SCA) has a geometry such that its comple-
ment (where air is replaced by metal and metal replaced by air) can exactly overlay
the original structure through translation and/or rotation [61]. The principle states
that such antennas have constant impedance independent of the frequency, and ap-
plies to any shape of the structure. Figure 1.11 illustrates rectangular and circular
shape SCAs. In theory, the shapes extend infinitely, however, in practice they are
truncated in size [61] as the shape extends from the feed point of the antenna.
1.4. UWB technology for indoor localization 23

F IGURE 1.11: Two self-complementary antennas of different shapes

(rectangular and spiral) and log-periodic antenna in 3D and planar
forms.

This principle led directly to the invention of the class of antennas called log-
periodic dipole antennas (figure 1.11), which is a modified folded up SCA, and has
high gains and relatively constant omni-directional radiation patterns and impedances
over bandwidths of 10:1 and more. The log-periodic antenna is a multi-resonant
antenna, this type of antennas consists of an arrangement of multiple narrowband
radiating elements, this creates a large array resonating at multiple frequencies. De-
spite their UWB characteristic, these antennas are not suitable for IR-UWB systems
because their phase centers are not fixed in frequency which leads to dispersion [61].

Rumsey’s principle
Rumsey’s principle [63] states that for an antenna to be frequency independent, its
shape needs to be specified only in terms of angles (figures 1.12, and 1.13). Examples
of such antennas are the balanced spiral antenna, the logarithmic spiral antenna and
conical spiral antenna [63, 64].

F IGURE 1.12: Antennas defined by angles: Spiral antenna, logarith-

mic spiral antenna and conical spiral antenna.

F IGURE 1.13: Spiral antenna with unbalanced excitation and ground

plane.

These classic frequency independent antennas may exhibit broadband frequency

performance however they still have drawbacks such as their physical dimensions
24 Chapter 1. Towards high-precision localization of objects

which tend to be significant as both Mushiake and Rumsey’s principles describe

shapes which theoretically extend infinitely to obtain the desired bandwidth.
Other classic broadband antennas include horn antennas and tapered slot antennas
(TSA) (figure 1.14) [61], they belong to the traveling-wave antennas type and provide
directional radiation. Their principal characteristic is having a gradual transition
structure (flaring waveguide shape) which helps radiating the waves efficiently into
space, they were introduced in [65, 66].

F IGURE 1.14: Horn antennas (rectangular and circular cross-

sections), TSA antennas of different shapes.

Horn antennas’ shape can be a pyramid with rectangular cross-section to be used

with rectangular waveguides, or a cone shape with circular cross-section to be used
with cylindrical waveguides. TSA antennas consist of a tapered slot carved into the
metallization on top of a dielectric substrate, depending on the etched slot shape,
TSA antennas are categorized in: linear taped slot antenna (LTSA), constant width
tapered slot antenna (CWTSA), broken linearly tapered slot antenna (BLTSA) and fi-
nally the well-known Vivaldi antenna or the exponentially or elliptically tapered slot
antenna. TSAs radiate unidirectionally in the plane of the substrate; their directiv-
ity increases with frequency and some antennas can achieve up to 10 dBi of gain [61].

In addition to frequency independent antennas and TSAs, broadband antennas

have also been derived as an evolution of popular monopole and basic dipole an-
tennas [61]. Examples of these antennas are the biconical dipole antenna (Lodge’s)
and its planar evolution the bowtie dipole antenna, represented in figure 1.15.

F IGURE 1.15: Evolution from classic dipole antenna to biconical

dipole and bowtie antennas, wire monopole antenna with ground
plane orthogonal to the radiating pole [61].
1.4. UWB technology for indoor localization 25

These antennas represent a modified dipole which has thicker arms the more
it is extended from the feeding point. The biconical antenna also evolved towards
the discone antenna, which instead has a single cone, and is more limited in the
bandwidth, however, presents a more stable antenna phase center compared to the
biconical version [61].
Single cone antenna could also be seen as the evolution of monopole antenna by con-
sidering the second pole as a conducting ground plane, which will resemble the wire
monopole configuration (figure 1.15). Until today, manufacturers lean towards the
UWB monopole antenna more than any other structure, and widely integrate these
in most UWB RTLSs, due to the numerous advantages they present such as: their
simple structure, low cost, broadband bandwidth, omni-directional radiation pat-
tern and simple matching capabilities. However, since they are not planar structures
due to their ground plane being perpendicular to the radiating element, this makes
their integrability occupy a large area [61]. Hence, research has moved towards
printed monopole antennas and different shapes microstrip designs printed on a
dielectric substrate were proposed, such as the disc [67], elliptic [68] and two-step-
rectangular monopole antenna [69] and other shape modified antenna and ground
notched designs to enhance the bandwidth (figure 1.16). These antennas are fed by
microstrip lines or coplanar waveguide transmission lines.

F IGURE 1.16: Planar monopole UWB antennas: two-step band-notch

monopole antenna, circular monopole antenna, hexagonal monopole
antenna, elliptical monopole antenna with truncated ground plane.

1.4.9 Requirements for UWB antennas in localization systems

Although the interest in developing and designing Ultra-wideband (UWB) anten-
nas has been growing since the FCC of the United States has allocated the 3.1 - 10.6
GHz frequency range to UWB technology and its applications, the performance re-
quirements are no longer the same for these antennas since the IEEE.802.15.4a has
declared the use of UWB channels of either 500 MHz or 1333 MHz instead of all
of the 7.5 GHz bandwidth. The requirements also enhanced since the apparition of
the first UWB real time locating system from Decawave in 2014 [70]; Naturally the
change in the design specifications and analysis metrics for these antennas has fol-
lowed. In this section, we describe how these requirements have changed and what
they have become.

The old requirements

In previous literature work, UWB antenna researchers demonstrated that pulse dis-
tortion and undesired phase response variations (non-linear response) are hindrances
that affect the performance of the IR-UWB RTLS. Thus, classic figures-of-merits de-
pending on frequency, such as: reflection coefficient of the antenna, radiation pat-
tern, and gain, not only are required to be analyzed for all the ultra-wide bandwidth
26 Chapter 1. Towards high-precision localization of objects

frequencies, but they are also no longer sufficient to evaluate UWB antennas’ re-
sponse and performance. Consequently, new metrics were introduced to reliably
characterize and predict the behavior of UWB antennas, such as: the system fidelity
factor (SFF) which indicates the amount of distortion occurred in the transmitted
UWB pulse was caused by the antenna [71], by comparing its shape between the
transmit and receive sides of the RTLS. Another metric that was introduced, is the
distance estimation error (DEE) which indicates the amount of bias (error) is affect-
ing the ranging estimates.

The recent requirements

Initially, as the IR-UWB systems considered the transmission of baseband ultra-short
electromagnetic pulses over a bandwidth up to 7.5 GHz, antenna designers focused
their research on developing techniques that allow the bandwidth enhancement of
antennas such that they can transmit sub-nanosecond pulses, with the minimum
level of pulse distortion possible. However, UWB developments have followed at
system and standardization levels, where the specifications replaced the extremely
short time pulses (sub-nanoseconds) by modulated pulses of bandwidths of 500
MHz or 1 GHz (mostly 500 MHz), that is: pulse widths of 2 ns or more. Although
the bandwidth of these channels is smaller, the benefits of UWB such as the ability
to resolve individual multipath signals and to perform precise ranging remain un-
changed. Consequently, UWB antennas are now required to cover only a relative
bandwidth less than 10 %.
Furthermore, with recent advances in UWB RTLS development aiming for mass pro-
duction of these radio chips in the industry, other physical and electrical constraints
are being paid attention to, such as: small size of the antenna and the study of their
integration environment to guarantee their behavior in their integration platform.
For this reason, standard antenna simulations and design in free-space and stand-
alone conditions are no longer sufficient to adhere to all performance expectations
of the designed IR UWB RTLS.

1.5 Objectives and developed axes of the thesis

The general objective is to propose improvement and evolution perspectives to UWB
real-time locating systems, specifically through the optimization of UWB antenna
designs. Among the objectives initially foreseen, the developed leads and works
carried out in the context of this thesis has focused on the following aspects :

• Adaptability to new UWB standard specifications.

• Enhancement of current RTLS ranging quality through the conception of opti-

mized UWB antennas.

• Introduction of a new feature: multi-standard UWB-LoRa joint localization.

BIBLIOGRAPHY 27

Bibliography
[1] D. Dardari, P. Closas, et P. M. Djuric, « Indoor Tracking: Theory, Methods, and
Technologies », IEEE Trans. Veh. Technol., vol. 64, no 4, p. 1263-1278, avr. 2015,
doi: 10.1109/TVT.2015.2403868.

[2] D. Dardari et al., « Chapter 5 - Innovative Signal Processing Techniques for

Wireless Positioning », in Satellite and Terrestrial Radio Positioning Techniques,
D. Dardari, E. Falletti, et M. Luise, Éd., Oxford: Academic Press, 2012, p.
207-315. doi: 10.1016/B978-0-12-382084-6.00005-2.

[3] S. Salem Al-Bawri, M. Tariqul Islam, M. Jit Singh, M. Faizal Jamlos, A. Narbu-
dowicz et al., "Rss-based indoor localization system with single base station,"
Computers, Materials and Continua, vol. 70, no.3, pp. 5437–5452, 2022.

[4] C. Esposito, M. Ficco, Deployment of RSS-Based Indoor Positioning Systems.

Int J Wireless Inf Networks 18, 224–242, 2011, https://doi.org/10.1007/s10776-
011-0131-7.

[5] W. K. Zegeye, S.B. Amsalu, Y. Astake and F. Moazzami, "WiFi RSS fingerprint-
ing indoor localization for mobile devices," in Proceedings of the 2016 IEEE 7th
Annual Ubiquitous Computing, Electronics and Mobile Communication Con-
ference (UEMCON), New York, NY, USA, 20–22 October 2016.

[6] S. Phimmasean, T. Chuenurajit, and P. Cherntonomwong, "Indoor localization

system based on fingerprint technique using RFID passive tag," in Proceedings
of the 2013 10th International Conference on Electrical Engineering/Electronics,
Computer, Telecommunications and Information Technology, Krabi, Thailand,
15–17 May 2013; pp. 1–6.

[7] F. Subhan, H. Hasbullah, A. Rozyyev, and S. T. Bakhsh, "Indoor positioning in

Bluetooth networks using fingerprinting and lateration approach," in Proceed-
ings of the International Conference on Information Science and Applications,
Jeju Island, Korea, 26–29 April 2011.

[8] C. Zhang, N. Qin, Y. Xue, and L. Yang, "Received Signal Strength-Based In-
door Localization Using Hierarchical Classification," in Sensors 2020, 20, 1067,
https://doi.org/10.3390/s20041067

[9] D. Dardari, A. Conti, U. Ferner, A. Giorgetti, et M. Z. Win, « Ranging With

Ultrawide Bandwidth Signals in Multipath Environments », Proc. IEEE, vol. 97,
no 2, p. 404-426, févr. 2009, doi: 10.1109/JPROC.2008.2008846.

[10] K. Haneda, R. Rudd, E. Vitucci, D. He, P. Kyösti, F. Tufvesson, S. Salous, Y.

Miao, W. Joseph, E. Tanghe, "Chapter 2 - Radio propagation modeling methods
and tools," in "Inclusive Radio Communications for 5G and Beyond," Academic
Press, 2021, Pages 7-48, ISBN 9780128205815, https://doi.org/10.1016/B978-0-
12-820581-5.00008-0.

[11] J. Dai, M. Wang, B. Wu, J. Shen, and X. Wang, "A Survey of Latest Wi-Fi As-
sisted Indoor Positioning on Different Principles," in Sensors 2023, 23, 7961.
https://doi.org/10.3390/s23187961
28 BIBLIOGRAPHY

[12] G. Shen, R. Zetik, H. Yan, O. Hirsch and R. S. Thomä, "Time of arrival esti-
mation for range-based localization in UWB sensor networks," 2010 IEEE In-
ternational Conference on Ultra-Wideband, Nanjing, China, 2010, pp. 1-4, doi:
10.1109/ICUWB.2010.5614041.

[13] R. M. Buehrer, and S. Venkatesh, "Fundamentals of Time-of-Arrival-Based Posi-

tion Locations," In "Handbook of Position Location" (eds S.A. Zekavat and R.M.
Buehrer), 2011, https://doi.org/10.1002/9781118104750.ch6.

[14] N. S. C. J, N. H. A. Wahab, N. Sunar, S. H. S. Ariffin, K. Y. Wong, et Y. Aun, «

Indoor Positioning System: A Review », Int. J. Adv. Comput. Sci. Appl., vol. 13,
no 6, 2022, doi: 10.14569/IJACSA.2022.0130659.

[15] J. A. del Peral-Rosado, R. Raulefs, J. A. Lopez-Salcedo, et G. Seco-Granados,

« Survey of Cellular Mobile Radio Localization Methods: From 1G to 5G
», IEEE Commun. Surv. Tutor., vol. 20, no 2, p. 1124-1148, 2018, doi:
10.1109/COMST.2017.2785181.

[16] J. Pospisil, R. Fujdiak, and K. Mikhaylov, "Investigation of the Performance of

TDoA-Based Localization Over LoRaWAN in Theory and Practice," in Sensors
2020, 20, 5464.

[17] I. Daramouskas, D. Mitroulias, I. Perikos, D. Paraskevas, V. Kapoulas, "Lo-

calization in LoRa Networks Based on Time Difference of Arrival," Springer:
Cham, Switzerland, 2022; pp. 130–143.

[18] J. J. Pérez-Solano, S. Ezpeleta, J. M. Claver, "Indoor localization using time dif-

ference of arrival with UWB signals and unsynchronized devices," Ad Hoc Net-
works, Volume 99, 2020, https://doi.org/10.1016/j.adhoc.2019.102067.

[19] V. Sipal, M. John, D. Neirynck, M. McLaughlin, et M. Ammann, « Advent of

practical UWB localization: (R)Evolution in UWB antenna research », in The 8th
European Conference on Antennas and Propagation (EuCAP 2014), The Hague,
Netherlands: IEEE, avr. 2014, p. 1561-1565. doi: 10.1109/EuCAP.2014.6902082.

[20] C. Falsi, D. Dardari, L. Mucchi, et M. Z. Win, « Time of Arrival Estimation for

UWB Localizers in Realistic Environments », EURASIP J. Adv. Signal Process.,
vol. 2006, no 1, p. 032082, déc. 2006, doi: 10.1155/ASP/2006/32082.

[21] Fira consortium. Accessed: Dec. 2023. [Online]. Available:

https://www.firaconsortium.org/

[22] G. Bacci, E. Falletti, C. Fernández-Prades, M. Luise, D. Margaria, et F. Zanier,

« Chapter 2 - Satellite-Based Navigation Systems », in Satellite and Terrestrial
Radio Positioning Techniques, D. Dardari, E. Falletti, et M. Luise, Éd., Oxford:
Academic Press, 2012, p. 25-74. doi: 10.1016/B978-0-12-382084-6.00002-7.

[23] R. Mautz, « Indoor positioning technologies », p. 1 Band, 2012, doi:

10.3929/ETHZ-A-007313554.

[24] C. S. Álvarez-Merino, H. Q. Luo-Chen, E. J. Khatib, et R. Barco, « WiFi FTM,

UWB and Cellular-Based Radio Fusion for Indoor Positioning », Sensors, vol.
21, no 21, p. 7020, oct. 2021, doi: 10.3390/s21217020.
BIBLIOGRAPHY 29

[25] L. Wang et M. Zawodniok, « RSSI-based localization in cellular net-

works », in 37th Annual IEEE Conference on Local Computer Networks
– Workshops, Clearwater, FL, USA: IEEE, oct. 2012, p. 820-826. doi:
10.1109/LCNW.2012.6424069.

[26] Y. You et C. Wu, « Indoor Positioning System With Cellular Network Assistance
Based on Received Signal Strength Indication of Beacon », IEEE Access, vol. 8,
p. 6691-6703, 2020, doi: 10.1109/ACCESS.2019.2963099.

[27] M. A. C. Duran et al., « Chapter 3 - Terrestrial Network-Based Positioning and

Navigation », in Satellite and Terrestrial Radio Positioning Techniques, D. Dar-
dari, E. Falletti, et M. Luise, Éd., Oxford: Academic Press, 2012, p. 75-153. doi:
10.1016/B978-0-12-382084-6.00003-9.

[28] M. Ghavami, L. B. Michael, et R. Kohno, "Ultra Wideband Signals

and Systems in Communication Engineering," 2nd ed. Wiley, 2007. doi:
10.1002/9780470060490.

[29] A. Guerra, F. Guidi, D. Dardari, et P. M. Djuric, « Near-field Tracking with Large

Antenna Arrays: Fundamental Limits and Practical Algorithms », IEEE Trans.
Signal Process., vol. 69, p. 5723-5738, 2021, doi: 10.1109/TSP.2021.3101696.

[30] LoRaWAN, « Why LoRaWAN is the logical choice for asset tracking connectiv-
ity », April 2020.

[31] T. W. Barrett, “History of Ultra WideBand (UWB) Radar and Communica-

tions: Pioneers and Innovators”, Progress In Electromagnetics Symposium 2000
(PIERS2000), Cambridge, MA, Jul. 2000.

[32] G. B. Giannakis, “Ultra-Wideband Communications, An Idea whose Time has

Come”, Signal Processing Magazine, IEEE, Vol. 21 , pp 26 – 54, Nov. 2004.

[33] H. F. Harmuth, “Transmission of Information by Orthogonal Functions”, 1sr

Edition, Springer, 1969.

[34] H.F. Harmuth, “Range-Doppler Resolution of Electromagnetic Walsh Waves in

Radar”, IEEE Trans. Electromagn. Compat., EMC-17, 106-111, 1975.

[35] H.F. Harmuth, “Synthetic aperture radar based on nonsinusoidal functions

pulse compression, contrast, resolution, and Doppler shift”, IEEE Transaction
on Electromagnetic Compatibility, EMC-21, pp. 40-49, 1979.

[36] H.F. Harmuth, “Nonsinusoidal waves for radar and radio communication.”
New York: academic, 1981.

[37] H.F. Harmuth, “Antennas and waveguide for nonsinusoidal waves”, New
York: academic, 1984.

[38] G.F. Ross, K.W. Robbins, “Base-band radiation and reception system”. U.S.
Patent 3,739,392, Jun. 1973.

[39] G.F. Ross, K.W. Robbins, “Narrow range-gate baseband receiver”. U.S. Patent
4,695,752, Sep. 1987.

[40] G.F. Ross, “Transmission and reception system for generating and receiving
base-band duration pulse signals for short base-band pulse communication sys-
tem”. U.S. Patent 3,728,632, Apr. 1973.
30 BIBLIOGRAPHY

[41] P. Van Etten, “The present technology of impulse radars,” Int. Radar Conf. Proc.
535-539, Oct. 1977.

[42] R. A. Scholtz, “Impulse Radio: How it works,” IEEE Communications Letters,

Vol. 2, NO. 1, Jan. 1998.

[43] R. A. Scholtz, “Wide Bandwidth Time-Hopping Spread-Spectrum Impulse Ra-

dio for Wireless Multiple-Access Communications”, Ultra IEEE Transactions
On Communications, Vol. 48, No. 4, Apr. 2000.

[44] R. A. Scholtz, “Characterization of Ultra-Wide Bandwidth Wireless Indoor

Channels”, a Communication Theoretic View, IEEE Journal On Selected Areas
In Communications, Vol. 20, NO. 9, Dec. 2002.

[45] M. Z. Win, “On the Robustness of Ultra Wide Bandwidth Signals in Dense Mul-
tipath Environments,” IEEE communications letters, Vol. 2, NO. 2, Feb. 1998.

[46] M. Z. Win, “On the Energy Capture of Ultrawide Bandwidth Signals in Dense
Multipath Environments”, IEEE Communications Letters, Vol. 2, No. 9, Sep.
1998.

[47] M. Z. Win, “Performance of Rake Reception in Dense Multipath Channels: Im-

plications of Spreading Bandwidth and Selection Diversity Order”, IEEE Jour-
nal On Selected Areas In Communications, Vol. 18, NO. 8, Aug. 2000.

[48] M. Z. Win, “Spectral Density of Random UWB Signals”, IEEE Communications

Letters, Vol. 6, NO. 12, Dec. 2002.

[49] M. Z. Win, “Spectral Density of Random Time-Hopping Spread Spectrum UWB

Signals with Uniform Timing Jitter”, Military Communications Conference Pro-
ceedings. MILCOM. vol.2, pp 1196 – 1200, 1999.

[50] M. Z. Win, “A Unified Spectral Analysis of Generalized Time-Hopping Spread-

Spectrum Signals in the Presence of Timing Jitter,” IEEE JOURNAL On Selected
Areas In Communications, Vol. 20, NO. 9, Dec. 2002.

[51] J. Y Lee and R. A. Scholtz, "Ranging in a dense multipath environment using

an UWB radio link," in IEEE Journal of Selected Areas Communication, 20(9),
pp1677–1683, 2002.

[52] FCC, “Revision of part 15 of the commission’s rules regarding Ultra-Wideband

transmission systems”, Federal Communications Commission, Washington,
D.C., Tech. Ref. ET Docket 98-153, Apr. 2002.

[53] S. Frattasi, F.D. Rosa, and D. Dardari, "Ultra-wideband Positioning and Track-
ing," In "Mobile Positioning and Tracking (eds S. Frattasi and F.D. Rosa),". 2007,
https://doi.org/10.1002/9781119068846.ch9

[54] D. J. Taylor, “Introduction to ultra wideband radar systems,” CRC Press, 1995.

[55] IEEE Standard for Low-Rate Wireless Networks, in IEEE Std 802.15.4-2015
(Revision of IEEE Std 802.15.4-2011) , vol., no., pp.1-709, 22 April 2016, doi:
10.1109/IEEESTD.2016.7460875.

[56] P. Pagani, F. T. Talom, P. Pajusco, B. Uguen, "Ultra-Wideband Radio Propagation

Channels: A Practical Approach,” Wiley, 2008.
BIBLIOGRAPHY 31

[57] C. E. Shannon, “Communication in the Presence of Noise,” Proceedings Of The

Ire, vol. 37, no. 1, pp. 10–21, Jan. 1949.

[58] L.Fullerton, “UWB Waveforms and Coding for Communications and Radar,”
Telesystems Conference, Proceedings. Vol.1., NTC, National, pp : 139 - 141,
Mar. 1991.

[59] ECC, “L’utilisation dans des conditions harmonisées du spectre radioélectrique

pour des équipements fonctionnant grâce à la technologie à bande ultra large,"
Journal Officiel de l’Union Européenne, Avr. 2009.

[60] Apple Nearby Interactions. Accessed: Apr. 2023. [Online]. Available:

https://developer.apple.com/documentation/nearbyinteraction

[61] Y. Duroc and A. Imran, ‘UWB Antennas: Design and Modeling’, Ultra Wide-
band. Sciyo, Aug. 17, 2010. doi: 10.5772/10002.

[62] C. A Balanis, "Antenna Theory: Analysis and Design," 4th Edition | Wiley, (con-
sulted July 18th, 2023).

[63] V. H. Rumsey, “Frequency independent antennas,” IRE International Conven-

tion Record Vol. 5, pp. 114-118, Mar. 1957.

[64] J. D Dyson, “The equiangular spiral antennas,” IRE on antennas and propaga-
tion, pp. 181-187, 1959.

[65] N. E. Lindenblad, “Wide band antenna,” in U.S Patent 2,239,724, 29, Apr. 1941.

[66] L. N. Brillouin, “Broad Band Antenna,” in U.S. ATENTE 2, p 454, 766, Nov.
1948.

[67] J. Liang, C. C. Chiau, X. Chen , C. G. Parini, “Printed circular disc monopole

antenna for ultra-wideband applications,” Electronics Letters, Vol. 40, No. 20 3
(1246-1248), , Sep. 2004.

[68] C. Y. Huang, W. C. Hsia, “Planar Elliptical Antenna for Ultra wideband Com-
munications,” Electronics Letters, Vol. 41, No. 6, 2 (296-297), Mar. 2005.

[69] S. H. Choi, J. K. Park, S. K. Kim, and J. Y. Park, "A new ultra-wideband an-
tenna for UWB applications,” Microwave and Optical Technology Letters, Vol.
40, No.5, pp. 399-401, Mar. 2004.

[70] Qorvo DWM1000 UWB transceiver. Accessed: March. 2023. [Online]. Available:
https://www.qorvo.com/products/p/DWM1000

[71] D. Coppens, A. Shahid, S. Lemey, B. V. H. C. Marshall, et E. De Poorter, «

An Overview of Ultra-WideBand (UWB) Standards and Organizations (IEEE
802.15.4, FiRa, Apple): Interoperability Aspects and Future Research Di-
rections », IEEE Access, vol. 10, p. 70219-70241, 2022, doi: 10.1109/AC-
CESS.2022.3187410.
 33

Chapter 2

Time-domain analysis of
time-based ranging with UWB
antennas

2.1 Introduction
Starting from the IEEE amendment 802.15.4a, the transmitted UWB signals were
changed from extremely wideband baseband pulses [1, 2] to modulated pulses with
bandwidths of 500 MHz, that is, pulse durations equal or superior to 2 ns [3]. Conse-
quently, from that point on, the efforts in the theoretical research in antenna design
no longer needed to focus on finding solutions to the problems of extending antenna
bandwidths to several GHz to cover the whole UWB frequency spectrum [4], nor on
ensuring no pulse distortion occurs after the signal is transmitted in the propagation
channel [5–8]. However, the fact that it is more easily feasible for antennas to ac-
commodate 500 MHz bandwidths instead of 7 GHz bandwidths does not mean that
research in the UWB antenna field has found a line end, instead, the focus of this re-
search is turned to other interests, for example, studying the integration of the UWB
antennas in real localization systems [9]. Indeed, just like in any other technology,
designing UWB antennas as a stand-alone component is not sufficient to ensure that
high accuracy localization will be achieved. To conduct these studies effectively, it
is crucial to accurately simulate the UWB system as it operates in practical settings.
This involves modeling the UWB signal as it is applied to the transmitting antenna,
and subsequently examining its behavior after the antenna emits it into the propa-
gation channel towards the receiver antenna. Therefore, this chapter presents a time
and frequency domain study of the UWB signal and antennas with a focus on local-
ization applications, by evaluating time-of-light performance for different metrics :
reader-tag distance, and azimuth and elevation angles.

2.2 Time-domain analysis principle

Time domain analysis refers to the study of the changes in the UWB signal after it
has been sent by the antenna through the propagation channel [10]. This definition
applies to any signal and not just UWB signals. Specifically, here, for applications
of localization, the study focuses on extracting the time-of-arrival of the signal as it
is sent to reception probes [10–12], and then on analyzing the changes in this time
depending on various factors. These factors are expected to influence the UWB sig-
nal [13]. Here we focus on the distortion of the signal, influence of azimuth and
elevation angles, and the distances at which reception probes could be placed.
34 Chapter 2. Time-domain analysis of time-based ranging with UWB antennas

2.2.1 UWB signal processing chain

Figure 2.1 illustrates a UWB transceiver as specified by the IEEE organization. Al-
though, the IEEE standards 802.15.4 and 802.15.4z only specify the transmitter oper-
ation [14], the receiver is also illustrated here, as it performs the inverse operations
of the transmitter.

F IGURE 2.1: UWB Transceiver chain composition.

The transmitter consists of two inputs, the Physical Header (PHR) and a Payload.
The PHR contains UWB specific information, such as datarate, preamble duration
and frame lengths. The transmission chain operates by the following steps:

• First, the PHR is encoded with a Hamming code using "single-error correcting
and double-error detecting" (SECDED) coding method.

• The payload is encoded / decoded with a (63, 55) Reed-Solomon code.

• The PHR and the payload are then encoded by a convolutional code.

• The bitstream is then converted to complex symbols by a symbol mapper.

• Next, the preamble insertion block selects a code sequence which is then spread
and repeated, as the receiver expects the input waveform to begin with the
preamble.

• The pulse shaper receives the output of symbol mapping and preamble inser-
tion, which are ternary symbols (-1,0,1). The ternary symbol sequence is then
applied as an input to a conformal filter (for example: butterworth filter or
root-raised cosine filter) to create pulses.

• After pulse shaping, the ultra-wideband signal is transmitted over the air us-
ing UWB antennas.

In the following, the chapter focuses on the output UWB signal from the pulse
shaper, its transmission from the UWB antenna and its propagation in the air until
reception.
2.2. Time-domain analysis principle 35

2.2.2 The IEEE 802.15.4a complying transmitted UWB signal

The amendment IEEE 802.15.4a specifies the conditions to which any time domain
signal fed to UWB antennas must conform [3, 14], in terms of shape and also of
power spectral density mask. This conformance is verified by the correlation of
the used signal with a signal specified. For UWB signals, this specified signal is
of type root-raised-cosine pulse (RRC) and recently, of duration 2 ns (that is, 500
MHz of bandwidth in the frequency domain), for most UWB defined channels. The
amendment specifies that the correlation of the user-defined UWB pulse with the
required RRC pulse must be at least of 0.8 (80 % correlation) or higher for the used
baseband pulse to be considered complying with the UWB PHY Layer of the IEEE
standard.
The standard states that the transmitted pulse shape p(t) shall be constrained by
the shape of its cross-correlation function with a standard reference pulse, r(t). The
normalized cross-correlation between two waveforms is defined as follows [14] :
Z +∞
1
ϕ(τ ) = p Re r (t).p∗ (t + τ )dt (2.1)
Er E p −∞

where Er and E p are the energies of r (t) and p(t), respectively, p* denotes the
complex conjugate of p, and, Re indicates that the real part is used. The reference
r (t) pulse used in the calculation of |ϕ(τ )| is a RRC pulse with a roll-off factor of
β = 0.5. Mathematically, the equation of a RRC pulse is as follows [14, 15] :
sin[(1− β) Tπtp ]
cos[(1 + β) πt
Tp ] + 4β Ttp
4β
r (t) = p (4βt 2
(2.2)
π Tp 1−( Tp )

where Tp is the required duration for a pulse operating in a UWB channel, for exam-
ple the duration in channel 9 is 2 ns.
Furthermore, in order for a high-rate pulse (HRP) UWB PHY transmitter to be com-
pliant with this standard, the transmitted pulse p(t) shall have a magnitude of the
cross-correlation function |ϕ(τ )| whose main lobe magnitude is greater than or equal
to 0.8 for a main lobe duration Tw of at least 0.5 ns for channel 9, and any sidelobe
magnitude shall be no greater than 0.3.
To test a pulse for compliance, let |ϕ(τ )| be the magnitude of the cross-correlation of
p(t) and r (t), and let τi , for i = 1,2, . . . , be a set of critical points as follows [14] :

d
|ϕ(τ )|τ =τi = 0 (2.3)
dτ
The maximum happens at point τmax , where the condition |ϕ(τmax )| ≥ |ϕ(τ )| is
verified for all values of τ. In addition, the second constraint on the value of side-
lobes may be stated mathematically as |ϕ(τi )| ≤ 0.3 for all τi values.

Figure 2.2 illustrates an example of a HRP UWB-compliant pulse, p(t) (left plot),
along with the root raised cosine reference pulse r (t) (middle plot) with Tp = 2.0 ns
and the magnitude of the cross-correlation |ϕ(τ )| (right plot). The pulse p(t) is an 8
order butterworth pulse with a 3 dB bandwidth of 500 MHz. The figure is intended
to show that this example pulse meets the requirements for compliance. Specifically,
the main lobe is above 0.8 for nearly 1 ns, and no sidelobe is greater than 0.3 (in this
case, the largest sidelobe peak is 0.2). The pulse p(t) is a compliant pulse for UWB
channels {0:3, 5:6, 8:10, 12:14}.
36 Chapter 2. Time-domain analysis of time-based ranging with UWB antennas

F IGURE 2.2: Example of a compliant pulse: butterworth pulse p(t)

(left), root raised cosine reference pulse r(t) (middle), and magnitude
of their cross-correlation |ϕ(τ )| (right) [14].

2.2.3 UWB PHY Compliant signal generation

Although the conformance verification is based on the baseband pulse of bandwidth
500 MHz, the signal fed to the antenna consists in this pulse upconverted to the cen-
ter frequency of a chosen UWB channel. The user-defined time-domain pulse signal
can be of any type, for example, it can be a Butterworth pulse, or it can also be a RRC
pulse (as long as its cross-correlation with the reference pulse is as specified by the
standard).
In this work, MATLAB software is used to generate the signal which will be subse-
quently fed to the UWB antenna.

First, for the convenience of easier conformance a RRC pulse is chosen, described
by equation 2.2. Then, this signal is upconverted to the frequency 7.9 GHz (center
frequency of channel 9) by the principle of RF mixing with a continuous cosine wave
oscillating at this frequency. That is, the RCC is used for amplitude modulation of
the cosine signal. For this, the two signals are mixed by the multiplication of the
time-domain RRC pulse with the 7.9 GHz cosine wave.
Besides, it is worth to note, that UWB channel 9 was chosen for its worldwide avail-
ability. The design of the antenna used in the following study was also focused on
this channel.

The objective is to create a UWB signal ready to be applied at the antenna input to
test its time-domain response through EM simulations. However, Matlab and most
signal processing softwares, operate with samples to create signals, even continuous
ones. In fact, this continuity can only be observed if the number of samples is high
2.3. Analysis of an antenna operating at one UWB channel 37

enough to make the signal appear continuous. Consequently, requirements such as

the Nyquist requirement [16] need to be considered in the construction of the signal
with such softwares.

Nyquist requirement
Following on the number of samples used to create a time-domain signal, not only
this number has to be high enough to be able to observe a correct signal, the sam-
ple rate (number of samples per second), must be at least equal or higher than the
Nyquist frequency to avoid aliasing [17–19] as it causes the overlapping of frequency
components of the signal.
Indeed, aliasing causes distortions to the signal that is being constructed from sam-
ples, making it differ from the original continuous signal. Consequently, to avoid
aliasing and its negative impacts on the signal generated, the requirement suggests
that the chosen sample rate, must be at least twice the frequency of the signal (here
7.9 GHz), to even approximately represent the signal. A better rule of thumb is that
this rate should be at least 10 samples per period of the signal. Thus, for a signal at
7.9 GHz, a sampling rate of ideally 79 GHz to obtain a relatively clean representa-
tion of the signal. This would mean that, for example, for a 16 ns of time span, the
number of samples required is:

Nsamples = (time span).(sample rate) = (16.10−9 )(79.109 ) = 1264 (2.4)

It is worth to note that in CST Microwave, the maximal number of samples that
can be processed for the excitation signal is 40000 samples.

2.3 Analysis of an antenna operating at one UWB channel

Figure 2.3 illustrates the steps of the analysis. The IEEE compliant UWB pulse signal
generated, is used as the excitation signal for the designed UWB patch antenna in
CST Microwave. Field probes are placed at different distances from the antenna and
play the role of receivers, that allows to assess the signal after it has left the antenna.

F IGURE 2.3: The steps of the time-domain analysis of UWB signals’

propagation.
38 Chapter 2. Time-domain analysis of time-based ranging with UWB antennas

2.3.1 Excitation signal generation and processing using Matlab

The pulse chosen is a RRC pulse for the maximum correlation with the RRC specified
by the standard. It was constructed using a time width of 2 ns, a roll off factor of 0.5
and with a time span of 16 ns. Then, this pulse is mixed with a continuous cosine
wave oscillating at 7.9 GHz for modulation. The resulting signal is constructed with
8000 samples which corresponds to a higher number of samples necessary to avoid
aliasing. Figure 2.4 below illustrates the subject signals.

F IGURE 2.4: Construction of the IEEE compliant UWB pulse by RF

mixing.

2.3.2 Antenna characteristics

Figure 2.5 presents the UWB antenna studied, and its characteristics described be-
low:

F IGURE 2.5: Studied UWB patch antenna structure.

• Bandwidth: 500 MHz (7.75 GHz – 8.25 GHz), operation on the worldwide
available UWB channel 9.

• Antenna physical shape: probe-capacitively-fed UWB patch antenna.

• Antenna radiation type: 180° directional .

2.3. Analysis of an antenna operating at one UWB channel 39

• Polarization types studied: first, linear polarization, then circular polarization

added to the patch antenna by the means of a branchline coupler enabling two
orthogonal capacitive feeds.

• Excitation signals studied: default CST excitation, then IEEE standard speci-
fied UWB excitation signal (RRC pulse of bandwidth 500 MHz).

Figures 2.6 and 2.7 show the reflection coefficient and radiation patterns of this
antenna, respectively.

F IGURE 2.6: Reflection coefficient of the UWB antenna studied.

F IGURE 2.7: Radiation patterns of the UWB antenna, (left): 2-

dimensional polar pattern, (right): 3-dimensional pattern.

2.3.3 EM simulation
The resulting signal is applied to the input port of the designed UWB probe-fed
patch antenna by simulations in CST Microwave. The objective is to analyze its
performance by extracting the time-of-arrival of the UWB signal to the field probes.
First, since the farfield is the field of interest in this study, the near-field distance D
limit is calculated. For a certain frequency, it can be obtained by [20] :

2 d2 f
D= (2.5)
c
40 Chapter 2. Time-domain analysis of time-based ranging with UWB antennas

With d the largest length of the antenna, f is the center frequency of the channel,
and c the light velocity. We have f = 7.9 GHz and d = 25 mm, thus:

2(0.025)2 7.9 109

D= = 0.0329 m = 33 mm (2.6)
3 108
Starting from the distance 33 mm the near field area ends and the far field area
begins. For the remaining study, probes are placed at a minimum distance of 68
mm, which represents more than twice the limit D calculated, to ensure being at the
farfield region.

Received E-farfield assessment

The radiated Electric field (farfield) from the antenna is observed with probes placed
at positions of coordinates ( x pn , y pn , z pn ), with n an integer representing the probe
number. The positions of these probes are varied and the following metrics are eval-
uated: the electric field waveform and magnitude in time and frequency domains,
and the time of arrival at each probe.

• Observation probes along the z-propagation axis:

In this part, the aim is to observe the effect of distance increasing between a
transmitting UWB antenna and receiver probes placed: along the propagation axis
z (such as they are aligned with the center of the radiating element), thus the probes
have coordinates (0, 0, z pn ) as shown in figure 2.8. The coordinate z pn varies by probe
and is chosen such that it is placed in the far field region of the antenna (z pn >
3.3 cm).

F IGURE 2.8: Receiver probes along the propagation axis of the an-
tenna, for E-field observation.

The probes were placed at 10 cm, 25 cm, 50 cm, 1 m, and 2 m on the propagation
axis. The resulting E-field magnitude in frequency domain for each receiving probe
is illustrated in the figure 2.9. It shows that, for all distances, the magnitude is high-
est at the antenna’s UWB channel frequencies (7.75 GHz – 8.25 GHz) and reaches its
maximum at the channel’s center frequency (7.9 GHz). Furthermore, the compari-
son between the E-fields at the different probes shows that the magnitude decreases
with increasing the distance at all the frequencies.
Figure 2.10 presents the resulting E-field in the time domain, at the same obser-
vation probes. The time signals observed are the magnitude (absolute value) of the
received E-field components time signals E( Ex , Ey , Ez ), which represents the abso-
lute value of the RRC shape at each probe. It can be seen that the RCC waveforms
do not present any distortions. Moreover, the important characteristic presented by
these shapes is their time of arrival, as it shows that, the further the probe is from the
2.3. Analysis of an antenna operating at one UWB channel 41

F IGURE 2.9: Received E-field waveform in frequency domain, and its

variation with distance between the antenna and the probes.

antenna, the later in time the signal arrives. This time of arrival at the probe is the
time used in ranging and localization to compute the distance between the reader
and the target.

F IGURE 2.10: Received E-field waveform in time domain, and its

variation with distance between the antenna and the probes.

• Observation probes along y-axis (in front of the antenna):

In this part, the purpose is to observe the effect of y-axis probes on the time of
arrival, in front of the antenna, where the target’s position can happen to be. Three
probes were placed: along the y axis, at the same z pn distance of 1 m, and the same
x pn=0 coordinate corresponding to the center of the antenna (figure 2.11). Thus, the
probes have coordinates (0, y pn , 1) and the coordinate y pn varies by probe.
First, for a clearer analysis, we analyze, the E-farfield components Ex , Ey , Ez , at
one probe placed on the axis of propagation at a 1 m distance, that is, at coordinates
(0,0,1). Figure 2.12 presents the resulting waveforms of the components of the field
in the time domain.
42 Chapter 2. Time-domain analysis of time-based ranging with UWB antennas

F IGURE 2.11: Receiver probes along y-axis of the antenna, for E-field
observation.

The figure shows that the resulting time of arrival of all the waveforms Ex (t), Ey (t),
Ez (t) is the same as expected, with a Ez (t) signal magnitude being null because the
field has no component on the direction of propagation.
Furthermore, a difference in magnitude is observed between Ex (t), Ey (t) at all time
instants, this difference is due to the excitation point of the antenna not being com-
pletely on the diagonal of the axes x and y, as it was not a necessity during the design
process. A further observation is that the signal waveforms observed are identical
to the input RRC waveform and present little to no distortion in shape. These ob-
servations remain the same for the waveforms of all E-farfield probes regardless of
their coordinates. In the following, we compare between the resulting E-farfields at
different probes’ using only one field component for all, that is, either Ex (t) or Ey (t).

F IGURE 2.12: Waveforms of the components of the field (Ex , Ey , Ez )

in the time domain, received at one probe.

We proceed with analyzing the farfield by its component Ex (t) on the y axis
at a distance of 1 m in front of the antenna, by placing the probes at coordinates
(0, y pn , 1), as shown in figure 2.11. The values of y pn simulated are 0 cm, - 6 cm, 6
cm , 10 cm, and -10 cm (with the reference 0 being the y coordinate of the center of
the radiating element of the antenna). Figure 2.13 presents the result of the received
Ex (t), and figure 2.14 presents a zoomed-in version on the shapes of the RRCs at the
center of the envelope.
The results show that the closer to the center of the antenna (y=0), the faster the
2.3. Analysis of an antenna operating at one UWB channel 43

F IGURE 2.13: E-field component waveform Ex (t) and its variation

with y-axis coordinates.

F IGURE 2.14: E-field component waveform Ex (t) and its variation

with y-axis coordinates, zoomed-in version.

field reaches the probe target (when analyzing for vertically placed probes on the y
axis, at the same z distance from the antenna and the same x coordinate), hence the
observed signal Ex (t) has a smaller time of arrival for the probe of y=0 than all the
other probes, and the signal that reaches last is the one corresponding to the probe
of y=10 cm. Furthermore, a symmetry around the center of the radiating element is
present, as the times of arrival are identical for probes placed symmetrically (y=6 cm
and y=- 6 cm) and (y=10 cm and y= -10 cm).
The difference in the time of arrival is expected as the probes are in the farfield re-
gion, where the antenna can be modeled by a source point. Thus, the time of arrival
measured corresponds to the time it takes the signal to travel from the center of the
antenna to each probe, which makes the probes at y > 0 or y < 0 farther from the
antenna compared to the probe at y = 0.
44 Chapter 2. Time-domain analysis of time-based ranging with UWB antennas

• Observation probes at the extremities (in front and at the sides of the antenna):

In the previous experiment with probes, it was found that the time of arrival is al-
most identical for probes at different height points, at the same distance in front of
the antenna. The objective of the following experiment, is to evaluate if this time of
arrival is also identical for probes at different points in around the antenna. For this,
three probes are placed as follows: the first in front of the antenna, and the other
two at both the left and right extremities (figure 2.15). The three probes are far away
from the antenna by the same distance, first 10 cm and then 1 m distances are tested.

F IGURE 2.15: Receiver probes around the antenna, for E-field obser-
vation.

First, figure 2.16 illustrates the received magnitude of the E-field at each of the six
probes. Results show that, for probes placed at the same distances from the antenna
(10 cm or 1 m), the magnitude is higher in the case of the probe placed in front of the
antenna compared to the extremities, in both distance cases 10 cm and 1 m. This is
expected as the gain of the patch antenna is the highest in front of the antenna and
weaker at the extremities.

F IGURE 2.16: Received E-field magnitude waveform in time domain,

and its variation with probes at center in front and at the extremities
of the antenna.

2.3.4 Discussion
The analysis of the time of arrival of the E-field radiated by the UWB antenna has
shown that:
2.4. Conclusion 45

• Probes (or any receiving tag) placed at a distance from the antenna will be
characterized by the same time-of-arrival of the UWB signal as long as they
belong to the circumference of the same sphere where the antenna is at the
center (figure 2.17). This time will thus, yield identical distance estimates. This
suggests that azimuth and elevation angles, at which a tag can be placed, do
not influence the time-of-arrival of the signal if two tags are both placed at an
azimuth (or elevation) circle of the same radius, around the antenna.

• The radiation pattern of an antenna does not have an influence on the accuracy
of localization using time-based techniques, such as time-way ranging.

F IGURE 2.17: A, B, C tags yield the same times of arrival.

2.4 Conclusion
In this chapter, a time-domain analysis was performed to study UWB antennas and
signals from transmission to reception.
An excitation signal UWB pulse conformal to the IEEE standard was generated and
applied at the input of a UWB antenna of 500 MHz of bandwidth. In the context
of localization with time-based technique, receive probes were placed at different
distances and positions around the antenna to study the radiated electric field shape
and the time of arrival of the signals. Results showed that the accuracy of local-
ization with UWB antennas, is not influenced by different azimuth and elevation
angles (same times of arrival) of the radiation pattern, however this latter influences
the magnitude level of the signal as the antenna gain varies depending on the direc-
tion studied.
46 BIBLIOGRAPHY

Bibliography
[1] K. Siwiak and D. McKeown, "Ultra Wideband Radio Technology", 2nd Edition,
chapter "Generating and transmitting UWB signals". Chichester: John Wiley
and Sons Ltd, 2004, doi: 10.1002/0470859334.

[2] V. Sipal, B. Allen, D. Edwards, B. Honary, "Twenty years of ultrawideband:

Opportunities and challenges," Communications, IET , vol.6, no.10, pp.1147-
1162, July 3 2012, doi: 10.1049/iet-com.2011.0281.

[3] "IEEE Standard for Local and metropolitan area networks–Part 15.4: Low-
Rate Wireless Personal Area Networks (LR-WPANs)," in IEEE Std 802.15.4-
2011 (Revision of IEEE Std 802.15.4-2006) , vol., no., pp.1-314, 5 Sept. 2011, doi:
10.1109/IEEESTD.2011.6012487.

[4] John, M. and Ammann, M.J. (2005), "Optimization of impedance bandwidth

for the printed rectangular monopole antenna," Microw. Opt. Technol. Lett., 47:
153-154. https://doi.org/10.1002/mop.21109

[5] M. John and M. J. Ammann, "Antenna Optimization With a Computation-

ally Efficient Multiobjective Evolutionary Algorithm," in IEEE Transactions
on Antennas and Propagation, vol. 57, no. 1, pp. 260-263, Jan. 2009, doi:
10.1109/TAP.2008.2009775.

[6] L. Lizzi, F. Viani, R. Azaro and A. Massa, "Optimization of a Spline-Shaped

UWB Antenna by PSO," in IEEE Antennas and Wireless Propagation Letters,
vol. 6, pp. 182-185, 2007, doi: 10.1109/LAWP.2007.894157.

[7] A. J. Kerkhoff and H. Ling, "Design of a Band-Notched Planar Monopole

Antenna Using Genetic Algorithm Optimization," in IEEE Transactions on
Antennas and Propagation, vol. 55, no. 3, pp. 604-610, March 2007, doi:
10.1109/TAP.2007.891563

[8] M. John and M. J. Ammann, "Wideband Printed Monopole Design Using a Ge-
netic Algorithm," in IEEE Antennas and Wireless Propagation Letters, vol. 6,
pp. 447-449, 2007, doi: 10.1109/LAWP.2007.891962.

[9] V. Sipal, M. John, D. Neirynck, M. McLaughlin and M. Ammann, "Advent of

practical UWB localization: (R)Evolution in UWB antenna research," The 8th
European Conference on Antennas and Propagation (EuCAP 2014), The Hague,
Netherlands, 2014, pp. 1561-1565, doi: 10.1109/EuCAP.2014.6902082.

[10] M. J. Ammann, A. Dumoulin, M. John and P. McEvoy, "Time-domain perfor-

mance of printed UWB antennas," 2009 3rd European Conference on Antennas
and Propagation, Berlin, Germany, 2009, pp. 3766-3769.

[11] L. Guo, J. Liang, C.C.Chiau, X. Chen, C.G.Parini and J. Yu, “ Performance of

UWB disc monopoles in time domain”, IEE Proceedings on Microwaves, An-
tennas and Propagation, vol.1, no.4, August 2007, pp. 955-959, doi: 10.1049/iet-
map:20050253.

[12] J. Liang, L. Guo, C. C. Chiau and X. Chen, "Time domain characteristics of UWB
disc monopole antennas," The European Conference on Wireless Technology,
2005., Paris, France, 2005, pp. 289-292, doi: 10.1109/ECWT.2005.1617714
BIBLIOGRAPHY 47

[13] A. Dumoulin, M. John, M. J. Ammann and P. McEvoy, "Performance evaluation

of antennas for UWB radar and positioning systems," IET Irish Signals and Sys-
tems Conference (ISSC 2009), Dublin, 2009, pp. 1-6, doi: 10.1049/cp.2009.1729.

[14] "IEEE Standard for Low-Rate Wireless Networks," in IEEE Std 802.15.4-2015
(Revision of IEEE Std 802.15.4-2011) , vol., no., pp.1-709, 22 April 2016, doi:
10.1109/IEEESTD.2016.7460875.

[15] D. Dardari, A. Conti, U. Ferner, A. Giorgetti and M. Z. Win, "Ranging With Ul-
trawide Bandwidth Signals in Multipath Environments," in Proceedings of the
IEEE, vol. 97, no. 2, pp. 404-426, Feb. 2009, doi: 10.1109/JPROC.2008.2008846.

[16] C. E. Shannon, "Communication in the Presence of Noise," in Proceedings of

the IRE, vol. 37, no. 1, pp. 10-21, Jan. 1949, doi: 10.1109/JRPROC.1949.232969.

[17] L. Yong, S. Lv, and X. Wang, "Adaptive Sub-Nyquist Spectrum Sensing

for Ultra-Wideband Communication Systems" Symmetry 11, no. 3: 342,
https://doi.org/10.3390/sym11030342.

[18] S. Cui, "Performance analysis of impulse-radio ultra-wideband techniques for

low-rate communications," Doctoral thesis, Nanyang Technological University,
Singapore, 2014, https://doi.org/10.32657/10356/61813.

[19] S. -W. M. Chen and R. W. Brodersen, "A Subsampling Radio Architecture for
Ultrawideband Communications," in IEEE Transactions on Signal Processing,
vol. 55, no. 10, pp. 5018-5031, Oct. 2007, doi: 10.1109/TSP.2007.896056.

[20] C. A. Balanis, "Antenna Theory: Analysis and Design (3rd ed.)", 2005, Chapter
2, page 34.
 49

Chapter 3

Advancements in Industrial RTLSs:

A Technical Review of UWB
Localization Devices Emphasizing
Antennas for Enhanced Positioning

3.1 Introduction
The internet of things paradigm has brought over the past few years, renewed sig-
nificant interest to indoor positioning, tracking and localization topics, principally
as the real-time locating technology allows a reference node to infer the position of
tagged target nodes, opening then the leading road to millions of object-to-object
awareness applications. This chapter presents a technical review of the available
industrial and commercial UWB real-time locating transceivers (RTLS) is realized,
with a focus on the frontend antennas integrated in these systems to establish the
needed wireless communication for positioning. Then, the different characteristics
of these antennas are summarized and discussed along with their impact on the lo-
calization performance, notably in terms of the reading range, position information
accuracy, object-orientation independent localization and multipath mitigation, and
solutions are suggested to bring an antenna-based improvement to the RTLSs’ per-
formance.

3.2 UWB Technology in Real-Time Locating Systems

The popularity of positioning and localization field in research has always been soar-
ing since the 1970s, as the US Department of Defense (DoD) decided to design and
launch a positioning service with global coverage and continuous-time availability,
the Global Positioning System (GPS) [1], at first limited for military use only and
then, a decade later, expanded for civil users. This popularity is mostly always re-
lated to space-based positioning performed by global navigation satellite systems
(GNSSs) such as GPS, Galileo, Glonass and BeiDou, and involving a constellation of
satellites fixed on earth-centered orbital planes. However, satellites perform poorly
when position-information is sought in harsh propagation environment areas, typi-
cally indoor (inside buildings, mines. . . ), where multipath signals reflected from ob-
jects and obstacles obscure the localization desired signal. To provide reliable posi-
tioning indoors, many terrestrial-based wireless communication systems have been
explored and equipped with localization abilities [2], such as localization through
WiFi and Bluetooth technologies [3] widely available nowadays in almost any con-
nected mobile device [4, 5], or through radio frequency identification (RFID) [6] by
 Chapter 3. Advancements in Industrial RTLSs: A Technical Review of UWB
50
Localization Devices Emphasizing Antennas for Enhanced Positioning

adjusting subtly its architecture to be able to provide position-information of an ob-

ject in addition to its identification [7]. However, as these technologies were not
initially designed to specifically provide localization services, most of them tend to
be limited in important aspects such as accuracy, space coverage and reading range
at which objects can still be detected. To tackle these issues, ultrawide bandwidth
(UWB) technology was introduced in radio transceivers [8] and evolved with IEEE
standards [9, 10], which are then referred to as real-time locating systems (RTLS)
and are able to draw consistent centimeter-level accuracy in positioning of objects
indoors through their high robustness against parasitic multipath components [11].
The purpose of this work is to study how antennas can improve a RTLS’s localization
performance through a practical approach. For this reason, the authors propose a
technical review of the commercially available industrial UWB radio chips and their
evaluation boards, and a highlight summary of the frontend antennas employed or
integrated in such systems. Furthermore, we discuss the error sources that can affect
the radio propagation of the UWB signal between a reader and a tag in different sce-
narios, line-of-sight (LOS) and non-line-of-sight (NLOS). These error sources affect
the detection of the desired UWB signal from which the distance is supposed to be
computed for ranging. The influence of the characteristics of the antennas on the
location estimates provided by an RTLS are then discussed and solutions are sug-
gested.
The remainder of this chapter is organized as follows: Section 3 is an overview of
the error sources in time-based localization for both LOS and NLOS scenarios, then,
Section 4 presents the technical review of today’s industrial UWB radio transceivers
and RTLSs from different manufacturers, followed by a focus and summary of the
antennas employed in each system. Moreover, Section 5 discusses and details the
influence of antennas on the UWB localization performance and distance estimates
in terms of reading range, accuracy, and tag-orientation dependent detectability and
the solutions for improvement. Finally, conclusions are presented in Section 6.

3.3 UWB Localization from an industry point of view

At the earliest stages of real-time locating systems (RTLSs), technologies started from
exploiting ultra-sound, sonar (in radar domains), laser from optics (such as teleme-
ters) and vision technologies (such as in the robotics domain) [12, 13]. However,
recently the current trend to approach tracking and localization problems is to use
standard, low cost and already deployed technologies. Consequently, in the follow-
ing, we focus only on technologies based on radio signal exchange.
Moreover, the indoor environments are very challenging; thus, multiple technolo-
gies have been proposed to cope with their constraints. Indeed, throughout the
years, indoor localization has been studied and performed using several technolo-
gies, each conforming to different standards and radio frequencies. These technolo-
gies are heterogenous and include notably Wi-Fi, Bluetooth, Radio-frequency IDen-
tification (RFID), Near Field Communications (NFC). Although, these wireless com-
munication standards have been designed for other purposes and are not intended
for localization and ranging purposes, this did not prevent their exploitation for po-
sitioning at the cost of some localization performance limitations such as shorter
range, low or average accuracy and, small detectability coverage. However, since
the emergence of UWB technology for the specific use in ranging, the industry fol-
lowed by producing UWB-based radio chips ready for deployment in localization
and positioning, that is for the measurements of the reader-to-object distance and
3.3. UWB Localization from an industry point of view 51

the direction angle of one or several objects. In this section, a review of the existing
UWB radio chips and their RTLSs infrastructures (evaluation boards) is realized, and
then a comparative focus on the antennas employed in such systems is performed.

3.3.1 Overview of industry available UWB chips and evaluation boards

Table 3.1 presents a summary of the UWB radio chips available in the industrial
market and that are aimed at customers that wish to localize objects inside any in-
door environment. The UWB spectrum frequencies these chips operate on and the
localization techniques they allow are summarized.

Manufacturer UWB chip Freq Loclization technique

channel
Qorvo DW1000 [14] 1, 2, 3, 4, Two-way ranging or
5 TDoA
Qorvo DW3000 [15] 5, 9 Two-way ranging or
TDoA
Apple U1 [16] 5, 9 Two-way ranging,
TDoA or AoA
NXP SR150 [17] / SR040 5, 9 Two-way ranging,
[18] TDoA or AoA
Sunway SW UWB chip [19] 5,9 Two-way ranging,
TDoA or AoA
STMicroelectronics MOD1 [20] 1,2,3,4 Two-way ranging or
TDoA

TABLE 3.1: Summary of the UWB radio chips available in the indus-
try, their frequencies and localization techniques.

In the following, most of the listed UWB chips and their characteristics are dis-
cussed in detail, notably their antenna type and aimed performances.

Qorvo DWM1000 chip

Qorvo’s DWM1000 module is based on the DW1000 UWB transceiver IC, which is an
IEEE 802.15.4a UWB implementation. It integrates antenna, all RF circuitry, power
management and clock circuitry in one module. It can be used in two-way ranging
or TDOA location systems to locate assets to a precision of 10 cm and supports data
rates of up to 6.8 Mbps.
The antenna used in the module is the Partron dielectric chip antenna (part num-
ber ACS5200HFAUWB [21]), which is a Chip Antenna covering UWB bands from
3.2 GHz to 7.2 GHz. The antenna is shown in figure 3.1 and its characteristics are
summarized in table 3.2.
 Chapter 3. Advancements in Industrial RTLSs: A Technical Review of UWB
52
Localization Devices Emphasizing Antennas for Enhanced Positioning

F IGURE 3.1: DWM1000 frontend monopole ceramic pcb UWB an-

tenna [21].

Parameter Value
Type Non ground mounting type monopole ceramic antenna
Peak gain (azimuth) 4.16 dBi at 6.2 GHz
Average gain (azimuth) -1.10dBi at 6.2 GHz
Dimensions 8.0 mm x 6.0 mm x 1.2 mm
Polarization Linear
Matching impedance 50 Ω

TABLE 3.2: Characteristics of the UWB monopole antenna integrated

in the Qorvo DWM1000 transceiver [21].

When designing the PCB onto which DWM1000 will be soldered, the proximity
of the DWM100 on-board ceramic monopole antenna to metal and other non-RF
transparent materials needs to be considered carefully. Two suggested placements
schemes are shown in figure 3.2. For the best RF performance, Qorvo advices that
the ground copper should be flooded in all areas of the application board, except
in the areas marked “Keep-Out Area” in the figure, where there should be no metal
either side, above or below, such as batteries.

F IGURE 3.2: DWM1000 application board keep-out area for UWB an-
tenna integration [14].

The antenna radiation pattern, was measured in an anechoic chamber for three
planes and results were reported by Qorvo [14]. As the antenna is linearly polarised,
the vertically polarised field (Theta) is measured in the azimuth plane (figure 3.3)
and the horizontally polarised field (Phi) is measured in the elevation planes 1 and
2 (figures 3.4 and 3.5 respectively).
3.3. UWB Localization from an industry point of view 53

F IGURE 3.3: DWM1000 antenna radiation pattern charateristic on az-

imuth plane [14].

F IGURE 3.4: DWM1000 antenna radiation pattern charateristic on el-

evation plane (1) [14].

F IGURE 3.5: DWM1000 antenna radiation pattern charateristic on el-

evation plane (2) [14].

From the figures, the monopole antenna has a dipole-like omnidirectional radia-
tion pattern which is beneficial for large coverage for tag detectability in localization.

• DWM1000 evaluation boards :

DWM1000 can be directly connected to an Arduino board to perform ranging. How-

ever, the latest version of the module called DWM3000 [15] is to be used with differ-
ent evaluation boards, which are the Qorvo’s DWM3000EVB [22] (figure 3.6) which
 Chapter 3. Advancements in Industrial RTLSs: A Technical Review of UWB
54
Localization Devices Emphasizing Antennas for Enhanced Positioning

is an Arduino form factor compatible shield designed for the evaluation of this mod-
ule for use in a scalable real time location system. Another evaluation board is the
Makerfab’s ESP32 UWB board [23]. In the latter, the module is integrated on the
same pcb as the microcontroller. It has two versions the UWB-ESP32 (figure 3.7) and
the UWB-ESP32-Pro (figure 3.8) which additionnally includes an OLED display for
the location information.

F IGURE 3.6: DWM3000 evaluation board as arduino shield [22].

F IGURE 3.7: ESP32 all-in-one evaluation board for UWB tracking [23].

F IGURE 3.8: ESP32-Pro (with display) all-in-one evaluation board for

UWB tracking [23].

STMicroelectronics’ MOD1 chip

Similarly to the DWM1000 and DWM3000 modules, the MOD1 module (figure 3.9)
is a surface mount module for UWB high precision developped by french company
previously called BeSpoon, which now is part of STMicroelectronics. It was de-
signed for warehousing and logistics in companies, personal navigation devices, se-
curity and building controls, robotics, wearable sensors, healthcare maintenance,
3.3. UWB Localization from an industry point of view 55

virtual reality (VR) and gaming. It supports Two-Way Ranging and achieves preci-
sions down to 10 cm, it has datarates of up to 2 Mb/s.
The module has its embedded API for firmware operation and updates and master-
slave mode selection. The MOD1 module itself does not include an antenna and is
only available when integrated in its evaluation kit which is presented in the follow-
ing.

F IGURE 3.9: STMicroelectronics UWB module MOD1 [20].

• MOD1 evaluation board :

MOD1 chip is integrated in the STMicroelectronics B-UWB-MEK1 evaluation board

[24] (figure 3.10) which is equipped with an STM32 microcontroller, it is can be used
as a fixed or as a mobile device. The use from four up to sixty boards is possible to
prototype a location system.

F IGURE 3.10: B-UWB-MEK1 application board for MOD1 module

with external antenna [24].

The MEK1 evaluation kit contains two identical transceivers and two UWB an-
tennas of type 3D monopole, its characteristics are summarized in table 3.3.
 Chapter 3. Advancements in Industrial RTLSs: A Technical Review of UWB
56
Localization Devices Emphasizing Antennas for Enhanced Positioning

Parameter Value
Frequency 2 - 4.77 GHz
S11 - 8 dB
Gain 2 dBi
Radiation Omnidirectional
Polarization Linear (Vertical)
Beamwidth (- 6 dB) 360° (azimuth) / 80° (elevation)
Impedance 50 Ω
Dimensions 44 mm x 67 mm x 62 mm

TABLE 3.3: Characteristics of the UWB 3D monopole antenna of the

STMicroelectronics B-UWB-MEK1 evaluation board [24].

Additionnally, the radiation patterns of the monopole antenna are shown in fig-
ure 3.11, for both the azimuth and elevation planes, and for all the center frequencies
the UWB channels 1, 2 and 3.

F IGURE 3.11: MEK1 UWB antenna radiation pattern charateristic on

azimuth and elevation planes [25].

From the figure, since the commercial antenna is linearly polarized (vertical po-
larization), it has E-plane (elevation) and H-plane radiations (azimuth). It can be
observed that the commercial antenna for channel 2 has omnidirectional radiation
along 360◦ with a gain of 2 dBi.

UWB chips interoperable with Apple’s U1 chip for interactions

Apple announced the availability of UWB modules and development kits for in-
teraction with its U1 chip. The modules are from either Qorvo or NXP companies.
First, the Qorvo DWM3000 previously described is compatible for interaction with
the U1 chip. Second, development kits that contain NXP’s Trimension chip (SR040
or SR150), they can be used by third party device developers. Since the DWM3000
architecture is similar to the DWM1000 described before, here, we focus on the
3.3. UWB Localization from an industry point of view 57

NXP based kits. These kits are commercialized by NXP’s partners including: Sun-
way Communications, Murata and Mobile Knowledge. They are approved by Ap-
ple® for the purpose of evaluating UWB-enabled accessories that leverage Apple’s
Nearby Interaction framework in order to interact with Apple products that include
the U1 chip. These modules and evaluation boards are listed in the following.

• Sunway Communications’s SW-EVK-2 board based on NXP SR150 chip :

Sunway Communications developped a FiRa [26] and IEEE 802.15.4z UWB compli-
ant wireless transceiver module [19] (figure 3.12) based on NXP’s SR150 SoC [17]. It
allows the location of objects in RTLSs to a secure range precision of 10 cm and an
angle precision of 3 degrees. It is characterized by the integration of the SR150 SoC,
antenna interface support, power management and clock control which simplifies
design integration with minimal RF design required (external antenna).
It supports TWR, TDoA or phase-difference of arrival (PDoA) for AoA scheme in a
variety of applications such as industrial, healthcare, smart home, smart retail RTLS,
secure hands-free payment and automotive access.
The module supports UWB channels 5, 6, 8, and 9 (frequency range from 6.24 to 8.24
GHz) and allows a support for a 3x1 element antenna array with JSC type antenna
connectors.

F IGURE 3.12: Sunway Communications’ UWB module SR150 inter-

face for ranging and direction finding [19].

The Sunway UWB Evaluation Kit (Sunway UWB EVK-2) [27] (figure 3.13) is com-
posed of two Sunway’s UWB modules in an anchor-tag configuration to allow locat-
ing the module operating as a tag and to build and evaluate a RTLS. This board is to
be mounted on NXP QN9090 board for operation.
 Chapter 3. Advancements in Industrial RTLSs: A Technical Review of UWB
58
Localization Devices Emphasizing Antennas for Enhanced Positioning

F IGURE 3.13: Sunway UWB module’s application board SW-EVK-2

with UWB antennas [27].

The antenna array [28] used in the Sunway evaluation kit is shown in Figure 3.14
and its characteristics are summarized in table 3.4.

F IGURE 3.14: Sunway Communications’ UWB pcb dual patch an-

tenna [28] for SW-EVK-2 evaluation board.

Parameter Value
Antenna type 2x1 dual pcb patch UWB antenna
Frequency UWB channel 5 (6.25 GHz - 6.75 GHz)
Reflection coefficient < - 10 dB
Peak gain 5.92 dBi
Beam width (3 dB) 120°
Impedance 50 Ω
Polarization Circular (at channel center frequency)
Antenna connectors 2 JSC-male with cables
Dimensions 51 mm x 40 mm x 2.0 mm

TABLE 3.4: Characteristics of the UWB dual patch antenna [28] of the
Sunway evaluation board [27].

From the information above, it is important to note that the Sunway module
(figure 3.12) [19] accomodates all UWB channels 5, 6, 8 and 9, however, the Sunway
3.3. UWB Localization from an industry point of view 59

dual antenna (figure 3.14) is only designed to accomodate UWB channel 5, thus if
this antenna is used in the evaluation kit, ranging may only be performed on channel
5.
Moreover, at the time of writing of this article, the radiation pattern characteristic
plot of the antenna was not published by Sunway.

Murata’s Type2BP-EVK board based on NXP SR150 chip

The Type2BP Development Kit by Murata (figure 3.15) [29], based on NXP SR150 en-
ables a wide range of IoT devices to perform localization tasks or create a setup with
multiple UWB anchors. The kit is based on Murata’s Type2BP [30] ultra-small UWB
module (figure 3.16.a), which includes clock, filters and peripheral components, as
well as NXP’s QN9090 Bluetooth LE solution and a USB-UART conversion IC. The
Type-2BP can be controlled through QN9090 with power being supplied via a UWB
cable or from the COM port of a PC. The kit integrates an on-board 3 element patch
UWB antenna array.

F IGURE 3.15: Murata UWB Trimension™ SR150 based application

board Type2BP EVK [29] with UWB antenna : (a) top view, (b) bottom
view, (c) perspective view.

In addition to the TYPE2BP module (figure 3.16.a) which is employed in the

Murata kit and based on NXP SR150 chip, the company also developped another
module called TYPE2DK (figure 3.16.b) which is instead based on the NXP SR040
module. also UWB modules which based on NXP UWB chipset. Both modules
integrates NXP UWB chipset, filters, clock, and peripheral components. These mod-
ules are ideally suited for general IoT devices, where TYPE2BP [30] is suited for the
reader and the TYPE2DK [31] is more suited for the tag role.
 Chapter 3. Advancements in Industrial RTLSs: A Technical Review of UWB
60
Localization Devices Emphasizing Antennas for Enhanced Positioning

F IGURE 3.16: Murata UWB NXP based modules : (a) Type2BP (built
upon the SR150 module), (b) Type2DK (built upon the SR40 module)

For further details, Murata TYPE2BP module is compliant with IEEE802.15.4z

HRP PHY, it supports TWR, TDoA and 2D and 3D AoA schemes. It supports UWB
channels 5 and 9, and data rates of 6.81Mbps, 7.8Mbps, 27.24Mbps and 31.2Mbps.
Its characteristics are summarized in table 3.5.

Parameter Value
Module type surface mount
Frequency UWB channel 5 and 9
Dimensions 6.6 mm x 5.8 mm x 1.2 mm
Antenna integration no

TABLE 3.5: Characteristics of Murata Type2BP module (based on

NXP SR150).

Conversely, the Type2DK module [31] based on NXP’s SR40 is an all-in-one UWB
+ Bluetooth LE combination module which integrates NXP QN9090 Bluetooth® LE
+ MCU chipset, and on-board antennas for UWB and Bluetooth. This module can
be used as a standalone tag which operates as it operates by coin-cell battery. It can
also be used with Murata Type2BP evaluation kit described previously to perform
ranging tests. Other characteristics are summarized in table 3.6.

Parameter Value
Module type surface mount
Frequency UWB channel 5 and 9
Dimensions 19.6 mm × 18.2 mm × 2.3 mm
Antenna integration UWB and Bluetooth LE

TABLE 3.6: Characteristics of Murata Type2DK module (based on

NXP SR040).

To conclude on Murata modules, the Type2BP module does not contain any an-
tenna. It contains the support (only) of 3 UWB antennas which in the evaluation
kit are the elements of a 3 patch type element antenna array UWB. While the Mu-
rata Type2DK has on-board integrated antennas, one for Bluetooth LE and the other
one for UWB of type printed monopole. Other specifications of the antennas are
provided by the manufacturer upon purchase.
3.3. UWB Localization from an industry point of view 61

3.3.2 Summary of the antennas present in industrial UWB chips and eval-
uation board
To conclude the technical review of UWB real-time located systems, with a focus on
antennas. Tables 3.7 and 3.8 present a summary of the UWB modules, their evalua-
tion boards and antennas. Table 3.7 highlights if the UWB chip itself or its available
evaluation boards integrate an antenna or not. Table 3.8 summarizes the character-
istics of the UWB antennas integrated to each UWB chip or evaluation board.

Manufacturer UWB Antenna integra- EV boards EV board an-

module tion tenna
Qorvo / De- DWM1000 Integrated ce- DWM3000EVB x
caWave or ramic monopole Arduino
DWM3000 Shield or Mak-
erfab’s ESP32
board
STMicroelec. MOD1 x MEK1 SMA-
connected
antenna
NXP SR150 x Murata *3 element
Type2BP-EVK pcb antenna
and Sunway array *3el-
SW-EVK-2 ement pcb
(works with antenna
separate NXP array with
MCU) CP
NXP SR40 integrated pcb Murata no addi-
monopole Type2DK- tional
EVK

TABLE 3.7: Summary of the industrial UWB chips, their evaluation

boards and their antenna integration.
 Chapter 3. Advancements in Industrial RTLSs: A Technical Review of UWB
62
Localization Devices Emphasizing Antennas for Enhanced Positioning

UWB module or Antenna Frequency Gain Radiation Polarization

Ev. board type (GHz) (dBi)
Qorvo DWM1000 pcb ce- 3.2 - 7.2 4.16 Omnidir. Linear
ramic
monopole
antenna
STMicroelec 3D 2 - 4.77 2 Omnidir. Linear
MEK1 ev. board Monopole
antenna
Murata Type2BP- 3 patch el- Ch5 and not pub- 180° dir. Linear
EVK (based on ement pcb Ch9 lished
NXP SR150 mod- antenna ar-
ule) ray
Sunway EVK-2 3 patch el- 6.25 - 6.75 5.92 180° dir. Circular
ev. board (based ement pcb GHz
on NXP SR150 antenna ar-
module) ray
Murata Type2DK pcb Ch5 and not pub- Omnidir. Linear
module (based monopole Ch9 lished
on NXP SR40 antenna
module)

TABLE 3.8: Summary of the UWB antenna characteristics of the in-

dustrial UWB chips and / or their evaluation boards.

3.4 Discussion of the antenna influence on the localization

performance
An attempt of ranging and localization of objects with electronic devices compris-
ing transmit and receive antennas, especially in time-based schemes which mostly
employ UWB technology, is a sensitive process as the final result depends greatly
on the nature of the captured signal at the antenna stage. Thus, frontend design
must be meticulous to detect the most correct signal, that is the shortest path signal,
which leads to the computation of the distance estimate that is closest to the real dis-
tance. The shortest path signal can be the direct path or the first arriving reflection
in case of complete blockage between the reader and tag). Consequently, the type
of antennas used at both the reader and tag matter significantly in a sense that they
can either improve or degrade the distance estimation depending on their charac-
teristics, such as: type of antenna, gain, polarization, radiation pattern, etc. In the
following, we discuss the influence antenna characteristics can have on the quality
of ranging results.

3.4.1 Influence on the localization range

Range of a wireless communication system is directly related to the received power
at the receiver side, in this case, at the reader side. This applies in all schemes, not
just in received-signal strength (RSS) based localization systems but also in time-
based systems such as UWB transceivers. Indeed, as the received power decreases
with increased distance between the reader and tag, it determines the limit distance
at which the transceiver can no longer detect the signal from the tag and thus can no
3.4. Discussion of the antenna influence on the localization performance 63

longer perform ranging.

Industrial manufacturers of UWB transceivers sometimes publish the measured per-
formance of their UWB chips and evaluation boards, such as Murata for example.
The performance in terms of reading range of their NXP SR150 chip based TYPE2BP
evaluation kit was measured and reported. Figure 3.17 presents Murata’s measure-
ment set up and environment, using two TYPE2BP boards integrating transmit and
receive antennas.

F IGURE 3.17: Range measurement set-up from Murata Type2BP eval-

uation board [29].

This measurement was aimed to infer the maximum possible reading range at
which the tag can be detected. Two Type2BP EVK are connected in a controller
(reader) and controlee (tag) configuration to check how long ranging is possible.
While doing this, results are reported with several conditions as stated by the manu-
facturer. These conditions correspond to the different possible antenna orientations
of the reader and tag relatively to eachother. The reported ranges are summarized
in the table 3.9 below.

EVK (responder) EVK (initiator) Max ranging distance

0° 0° Around 50 m
0° 90° 26 m
0° 180° 25 m
90° 90° 10 m
90° 180° 9m
180° 180° 4.5 m

TABLE 3.9: Range measurement results from Murata Type2BP evalu-

ation board [29].

The evaluation kit uses patch type antennas with a theoretical 180° directional
radiation, and with linear polarization. Consequently, the best case scenario is the
condition 0° orientation for the responder and 0° orientation for the initiator, which
is reported to achieve a maximum reading range of 50 m. Linear polarization affects
the other cases in the received power as cross-polarization (0°-90°, 90°-0°, 0°-180°,
90°-180°) or induces power loss [32, 33]. Moreover, even in the cases where there is
no cross-polarization, the radiation pattern also affects the detectability, especially at
longer distances, in the cases where the antennas are not facing eachother (such as
in 0°-180°, 90°-90° and 180°-180°). For this reason, a solution would be design omni-
directional reader and tag antennas to unsure the transceivers don’t miss eatchother
in the search phase and discovery process during two-way ranging no matter how
 Chapter 3. Advancements in Industrial RTLSs: A Technical Review of UWB
64
Localization Devices Emphasizing Antennas for Enhanced Positioning

they are placed relatively to eachother. These observations show the significant ef-
fect of the transmit and receive antennas on the performance of the RTLS in terms of
range. This matter is discussed in further details in Chapters 4 and 5.

3.4.2 Influence on the tag-orientation dependence

From the previous table 3.9, the worst case scenario is the when the responder and
initiator antennas are in a 180°-180° orientation configuration relatively to eachother.
It can be seen that the maximum distance at which the reponder is detected is 4.5
m, which means further from this distance, not only ranging is impossible but its
presence itself cannot be detected. Taking into account the application of UWB
transceivers, these are mainly aimed for indoor environments with several mobile
objects, unlike for example RFID tracing or Bluetooth tracing, a range of 4.5 m is
too low for UWB deployment areas (hospitals, airports, construction mines etc..).
Furthermore, in such areas tags and objects are mobile and their orientation cannot
manually be adjusted to face the reader’s antenna nor to match its vertical or hor-
izontal polarization. For this reason, solutions such as circular polarization of the
reader antennas [32, 34] are advised to mitigate cross-polarization induced power
loss, even while the tag’s polarization may be kept linear for design simplicity. This
matter is discussed in further detail in Chapters 4 and 5 of this thesis.

3.4.3 Influence on the accuracy

After the reading range, the performance in terms of accuracy of the ranging esti-
mates of the NXP SR150 chip based TYPE2BP evaluation kit was similarly measured
and reported. Figure 3.18 presents Murata’s measurement set up and results, using
two TYPE2BP boards.

F IGURE 3.18: Accuracy measurement from Murata Type2BP evalua-

tion board [29].

For this measurement, their condition on the antennas is that they are placed fac-
ing eachother. In most cases, conclusions on accuracy is statistical or based on many
measured samples by ranging and comparing to the real distance. The results table
shows that three samples A, B and C of ranging were taken at both channel 5 (cen-
tered at 6.49 GHz) and channel 9 (centered at 7.9 GHz) and averaged to one value at
the end, the values are reported in cm.
Taking into account the real distance of 2 m and the reported average detected dis-
tance at both chanels of 1.99 m, an accuracy of 99 % which is most likely due to the
fine temporal resolution of UWB signal itself and the placement of antennas in com-
plete line-of-sight without any blockage nor reflecting objects and at a distance of
only 2 m. These observations show that, in fact, the theoretical UWB accuracy level
3.5. Conclusions 65

(that is under 10 cm) can be achieved, however only in adjusted conditions as seen in
the previous cases. Accuracy depends heavily on the environment, and for practical
applications, these environments cannot be prealably decided. This calls for the de-
sign of antennas robust to reflections and multipath. For this reason, a solution was
proposed in our work [32], and consists in mitigating multipath signals with circular
polarization filtering. This method requires both the reader and tag antenna to have
circular polarization of the same sense (either right-hand or left-hand), which can
filter out reflections on metal as these will switch the CP sense in the middle of the
path before arrival. This matter is discussed in further detail in Chapter 5.

3.5 Conclusions
This chapter presented a study of UWB locating systems by, first highlighting the
error sources in indoor environments and that can affect time-based localization.
Then, a technical review of the currently available industrial UWB chips and real-
time locating systems used for object monitoring in indoor scenarios and complex
environments was realized. Besides, the antennas employed in the listed systems
and their characteristics were highlighted and summarized. Furthermore, a discus-
sion on the influence of frontend antennas of UWB transceivers on the localization
quality in terms of reading range, accuracy and object orientation-independent de-
tectability was discussed and solutions were suggested.
66 BIBLIOGRAPHY

Bibliography
[1] G. Bacci, E. Falletti, C. Fernández-Prades, M. Luise, D. Margaria, et F. Zanier,
« Chapter 2 - Satellite-Based Navigation Systems », in Satellite and Terrestrial
Radio Positioning Techniques, D. Dardari, E. Falletti, et M. Luise, Éd., Oxford:
Academic Press, 2012, p. 25-74. doi: 10.1016/B978-0-12-382084-6.00002-7.

[2] M. A. C. Duran et al., « Chapter 3 - Terrestrial Network-Based Positioning and

Navigation », in Satellite and Terrestrial Radio Positioning Techniques, D. Dar-
dari, E. Falletti, et M. Luise, Éd., Oxford: Academic Press, 2012, p. 75-153. doi:
10.1016/B978-0-12-382084-6.00003-9.

[3] C. S. Álvarez-Merino, H. Q. Luo-Chen, E. J. Khatib, et R. Barco, « WiFi FTM,

UWB and Cellular-Based Radio Fusion for Indoor Positioning », Sensors, vol.
21, no 21, p. 7020, oct. 2021, doi: 10.3390/s21217020.

[4] L. Wang et M. Zawodniok, « RSSI-based localization in cellular net-

works », in 37th Annual IEEE Conference on Local Computer Networks
– Workshops, Clearwater, FL, USA: IEEE, oct. 2012, p. 820-826. doi:
10.1109/LCNW.2012.6424069.

[5] Y. You et C. Wu, « Indoor Positioning System With Cellular Network Assistance
Based on Received Signal Strength Indication of Beacon », IEEE Access, vol. 8,
p. 6691-6703, 2020, doi: 10.1109/ACCESS.2019.2963099.

[6] T. Sanpechuda and L. Kovavisaruch, “A review of RFID localization: Applica-

tions and techniques,” in Proc. 5th Int. Conf. Electr. Eng./Electron., Comput.,
Telecommun. Inf. Technol., May 2008, pp. 769–772.

[7] K. Finkenzeller, RFID Handbook: Fundamentals and Applications in Contact-

less Smart Cards and Identification, 3rd ed. Hoboken, NJ, USA: Wiley, 2010.

[8] L. Yang, G. B. Giannakis, « Ultra-wide band communications, an idea whose

time has come », IEEE Signal Processing Magazine, vol4, p. 1053-5888, Novem-
ber 2004.

[9] IEEE Standard for Low-Rate Wireless Networks, IEEE Standard 802.15.4- 2020
(Revision of IEEE Standard 802.15.4-2015), Jul. 2020.

[10] IEEE Standard for Low-Rate Wireless Networks–Amendment 1: Enhanced Ul-

tra Wideband (UWB) Physical Layers (PHYs) and Associated Ranging Tech-
niques, IEEE Standard 802.15.4z-2020 (Amendment to IEEE Standard 802.15.4-
2020), Aug. 2020.

[11] D. Dardari, A. Conti, U. Ferner, A. Giorgetti, et M. Z. Win, « Ranging With

Ultrawide Bandwidth Signals in Multipath Environments », Proc. IEEE, vol. 97,
no 2, p. 404-426, févr. 2009, doi: 10.1109/JPROC.2008.2008846.

[12] D. Dardari, P. Closas, et P. M. Djuric, « Indoor Tracking: Theory, Methods, and

Technologies », IEEE Trans. Veh. Technol., vol. 64, no 4, p. 1263-1278, avr. 2015,
doi: 10.1109/TVT.2015.2403868.

[13] N. S. C. J, N. H. A. Wahab, N. Sunar, S. H. S. Ariffin, K. Y. Wong, et Y. Aun, «

Indoor Positioning System: A Review », Int. J. Adv. Comput. Sci. Appl., vol. 13,
no 6, 2022, doi: 10.14569/IJACSA.2022.0130659.
BIBLIOGRAPHY 67

[14] Qorvo DWM1000 UWB transceiver. Accessed: Dec. 2023. [Online]. Available:
https://www.qorvo.com/products/p/DWM1000

[15] Qorvo DWM3000 UWB transceiver. Accessed: Dec. 2023. [Online]. Available:
https://www.qorvo.com/products/p/DWM3000

[16] Apple Nearby Interactions, "Locate and interact with nearby devices using
identifiers, distance, and direction," Accessed: Dec. 2023. [Online]. Available:
https://developer.apple.com/documentation/nearbyinteraction/

[17] NXP Trimension SR150 UWB transceiver. Accessed: Apr. 2023. [Online]. Avail-
able: https://www.nxp.com/products/wireless-connectivity/trimension-
uwb/trimension-sr150-secure-uwb-solution-for-iot-devices:SR150

[18] NXP Trimension SR040 UWB transceiver. Accessed: Apr. 2023. [Online]. Avail-
able: https://www.nxp.com/products/wireless-connectivity/trimension-
uwb/trimension-sr040-reliable-uwb-solution-for-iot:SR040

[19] Sunway Communications UWB transceiver for ranging and direc-

tion finding. Accessed: Dec. 2023. [Online]. Available: https://en.sz-
sunway.com/product/wireless/RF-Connectivity-Module/383.html

[20] STMicroelectronics B-UWB-MOD1 Ultra-wideband module for high-

precision indoor location. Accessed: Dec. 2023. [Online]. Available:
https://www.st.com/en/wireless-connectivity/b-uwb-mod1.html?rt=db&
id=DB4392

[21] Abracon chip antenna integrated in Qorvo DWM1000 module. Accessed: Dec.
2023. [Online]. Available: https://abracon.com/chip-antenna/ACA-107-T.pdf

[22] Qorvo DWM3000EVB for DWM3000 module. Accessed: Dec. 2023. [Online].
Available:
https://www.qorvo.com/products/p/DWM3000EVB

[23] Makerfab ESP32 ev.board for DWM3000. Accessed: Dec. 2023. [Online]. Avail-
able: https://www.makerfabs.com/esp32-uwb-ultra-wideband.html

[24] STMicroelectronics B-UWB-MEK1 evaluation board for UWB MOD1 module.

Accessed: Dec. 2023. [Online]. Available: https://www.st.com/en/wireless-
connectivity/b-uwb-mek1.html

[25] STMicroelectronics BeSpoon UWB Omnidirectional Antenna BSP-UWBA-

Om4G, Datasheet V.1.0, BeSpoon, France, 2021.

[26] FiRa Consortium for UWB tracking. Accessed: Jan. 2024. [Online]. Available:
https://www.firaconsortium.org/

[27] Sunway Communications UWB evaluation board based on NXP

SR150. Accessed: Jan. 2024. [Online]. Available: https://en.sz-
sunway.com/product/wireless/RF-Connectivity-Module/385.html

[28] Sunway Communications UWB antenna. Accessed: Apr. 2023. [Online].

Available: https://en.sz-sunway.com/product/wireless/RF-Connectivity-
Module/411.html
68 BIBLIOGRAPHY

[29] Murata Type2BP UWB development board based on NXP SR150. Accessed:
March. 2023. [Online]. Available: https://www.nxp.com/products/wireless-
connectivity/trimension-uwb/murata-type2bp-evk-trimension-sr150-uwb-
development-kit:UWB-DEV-TYPE2BP-EVK

[30] Murata Type2BP UWB module based on NXP SR150. Accessed:

March. 2023. [Online]. Available: https://www.murata.com/en-
eu/products/connectivitymodule/ultra-wide-band/nxp/type2bp

[31] Murata Type2DK UWB module based on NXP SR040. Accessed:

March. 2023. [Online]. Available: https://www.murata.com/en-
eu/products/connectivitymodule/ultra-wide-band/nxp/type2dk

[32] A. Benouakta, F. Ferrero, L. Lizzi and R. Staraj, "Antenna Characteristics Con-

tributions to the Improvement of UWB Real-Time Locating Systems’ Reading
Range and Multipath Mitigation," in IEEE Access, vol. 11, pp. 71449-71458,
2023, doi: 10.1109/ACCESS.2023.3294622.

[33] A. Benouakta, F. Ferrero, L. Lizzi, L. Brochier and R. Staraj, "Measure-

ments of antenna polarization effects on Ultra-Wideband monitoring and
localization," 2021 IEEE Conference on Antenna Measurements and Ap-
plications (CAMA), Antibes Juan-les-Pins, France, 2021, pp. 589-590, doi:
10.1109/CAMA49227.2021.9703597.

[34] A. Benouakta, F. Ferrero, L. Lizzi and R. Staraj, "Frequency Reconfigurable and

Circularly Polarized Patch Antenna Over Dual Ultra-wideband Channels," 2022
16th European Conference on Antennas and Propagation (EuCAP), Madrid,
Spain, 2022, pp. 1-5, doi: 10.23919/EuCAP53622.2022.9769006.
 69

Chapter 4

Localization with UWB antennas:

the requirements of RTLSs

4.1 Introduction
As one of the typical and modern applications of UWB technology is the localization
and positioning of tags associated to mobile objects, commercial companies and sup-
pliers have and continue to release on the market, many boards, electronic modules,
and kits meant for localization. These solutions are pervasive and somehow very
profitable as they are popular within the industrials that constantly need to locate
assets (from small objects, to smartphones, to cars, etc.. . . ).
Accordingly, the growing mobility of these objects in wireless real-life scenarios
raises the expectations for the UWB antennas used, more so for the reader (or “an-
chor”) antennas. In fact, to localize an object associated to an UWB transmitting
tag-antenna, the UWB receiving anchor-antenna at the other side is expected to pos-
sess certain important characteristics.
Hence, this chapter investigates how to improve the quality of the location estimates
obtained with UWB localization devices, through choosing suitable antenna specifi-
cations. Moreover, it presents and describes the design of a frequency reconfigurable
and circularly polarized UWB patch antenna along two UWB channels of 500 MHz
meant for UWB anchors.

4.2 Frequency reconfigurable and circularly polarized patch

antenna over dual Ultra-wideband channels
This section describes the design and characterization of an UWB patch antenna
which is aimed to be an Anchor-antenna in UWB localization systems. It can oper-
ate distinctively in two different UWB frequency channels. Both of these channels
have 500 MHz bandwidth, the lowband channel is centered at 4 GHz and the high-
band channel at 6.49 GHz. Reconfigurability to switch from one channel to another
is achieved by changing the size of the radiating element. Additionally, this UWB
antenna is aimed to be circularly polarized over both channels, which is realized
by the means of two branchline couplers placed on a bottom layer. Results of re-
configurability and circular polarization operation are presented by simulation and
measurements.

4.2.1 Antenna design specifications

The interest in developping and using UWB antennas has been growing since the
FCC of the United States has allocated the 3.1 - 10.6 GHz frequency range to UWB
70 Chapter 4. Localization with UWB antennas: the requirements of RTLSs

technology and its applications [1]. UWB communication systems transmit and re-
ceive ultra-short electromagnetic pulses over a bandwidth of several GHz [2]. Such
systems call for optimized RF front-ends including their UWB antennas.
The most typical application of UWB technology is the localization and position-
ing of tags associated to mobile objects. Accordingly, the growing mobility of these
objects in wireless real life scenarios raises the expectations for the UWB antennas
used, more so for the "Anchor" (or reader) antennas. In fact, to localize an object
associated to an UWB transmitting tag-antenna, the UWB receiving anchor-antenna
at the other side is expected to possess certain important characteristics.
First, increasing the UWB localization system’s detection range raises the need for
the anchor antenna to have enough gain in order to be able to detect and localize the
tag at longer distances.
Second, improving the ability of localization raises the need for an ideally constant
received power at the anchor antenna side, no matter the relative orientations of
the anchor and tag antennas to each other [3]. This specification calls for circular
polarization (CP) at the anchor side in order to avoid the anchor-tag polarization
mismatch.
Hence, antennas on the anchor side of UWB localization systems, should not only
have an ultra-wide bandwidth, but also a higher gain and ideally, CP to be able to
overcome the problem of limited range and polarisation-dependent received power.
Moreover, these high gain and CP characteristics could be more easily achieved by
UWB antennas if the expectation for them to operate in all the UWB frequency range
is removed. In fact, the generally used UWB modules for tracking and localization
systems such as the DWM1000 from Decawave [4], operate in channels of 0.5 or 1.3
GHz of bandwidth only instead of all the UWB range.
However, already existing solutions for UWB anchor-antennas mostly represent an-
tennas that cover all the 7.5 GHz of UWB frequency range [5]. Similarly, in terms of
frequency reconfigurability, most existing solutions have the characteristic of being
"band-notch", as they all mostly tend to cover all the UWB frequency range (3.1-10.6
GHz) as a general case and then obtain the other small configured bands by reject-
ing the undesired parts of the spectrum. For example, in [6] the authors proposed
an 8-states frequency reconfigurable UWB monopole antenna with band rejections
at the C band, WLAN, and X band (figure 4.1).

F IGURE 4.1: 8-States frequency reconfigurable UWB monopole an-

tenna [6], with band rejections at the C band, WLAN, and X band by
two C-slots and inverted U-slot
4.2. Frequency reconfigurable and circularly polarized patch antenna over dual
71
Ultra-wideband channels

Also, in [7], a reconfigurable ground plane with a slot was used in order to ob-
tain switchable bands between, the whole UWB range, a low, middle and upper
UWB sub-bands. Also, in [8], all the UWB range and different sub-bands were ob-
tained in a Vivaldi antenna by using switches on its wings (figures 4.2 and 4.3).

F IGURE 4.2: UWB frequency reconfigurable antipodal Vivaldi an-

tenna [8]: a) operation on UWB mode, b) operation on seven switch-
able sub-bands.

F IGURE 4.3: Reflection coefficient of the UWB frequency reconfig-

urable antipodal Vivaldi antenna: a) operation on UWB mode and 3
sub-bands, b) operation on 7 switchable sub-bands [8].

Conversely, in this work, the realized anchor antenna combines the high gain
of patch antennas with an ultra wide bandwidth while focusing on each 500 MHz
UWB channel at a time, by respectively using capacitive feeding of the radiating el-
ement and by making the antenna frequency reconfigurable. Its reconfigurability in
frequency is based on the modification of the size of the patch and eventually by us-
ing three p-i-n diodes. Additionally, this antenna exploits the simple structure and
72 Chapter 4. Localization with UWB antennas: the requirements of RTLSs

symmetry of a patch to achieve a CP operation covering all the frequencies in an

UWB channel.
First, the design method of the UWB anchor antenna based on capacitive feeding is
described, followed by the simulation results in terms of impedance matching band-
width and CP operation. Then, the fabrication and experimental characterization of
the antenna is illustrated and compared to the simulation results. Finally, these re-
sults are discussed and followed by a general conclusion.

4.2.2 Anchor antenna design principle

The realized Anchor UWB antenna (figures 4.4, 4.5, and 4.6) is based on the standard
microstrip probe-fed patch antenna model. However, to increase the usually narrow
band of patch antennas, an air gap is introduced between the substrate and ground
plane. Furthermore, to compensate for the inductive effect introduced due to the
probe, a capacitive feeding mechanism is applied following the approach used in [4].
This capacitive feed is made of a probe-fed small patch placed on the same substrate
layer as the radiating element and at a gap distance from it. The dielectric substrate
is suspended in air as illustrated in figure 4.4. The anchor antenna is designed to
operate in two UWB frequency channels. From 3.75 to 4.25 GHz and from 6.25 to
6.75 GHz. The first band is called ”Channel 2” and the second band ”Channel 5”,
as used in the existing UWB localization system DWM1000 by Decawave [5] for
example.

F IGURE 4.4: Side view of the UWB anchor antenna.

F IGURE 4.5: Top view of the UWB anchor antenna.

The design steps of the UWB Anchor patch antenna are as follows :

• First, two UWB patch antennas operating along the frequency range of chan-
nel 2 and channel 5 (each has 500 MHz of bandwidth), are both designed sep-
arately.
4.2. Frequency reconfigurable and circularly polarized patch antenna over dual
73
Ultra-wideband channels

F IGURE 4.6: Bottom view of the UWB anchor antenna.

• Second, as in figure 4.5, both of the realized antennas are ”combined” into
one reconfigurable patch antenna by modifying the shape of the radiating ele-
ment using three p-i-n diodes. If the diodes are forward biased, the radiating
element operates in the low-band channel, and if the diodes are open, the ra-
diating element operates in the high-band channel.

• Finally, as in figure 4.6, CP is introduced to the frequency reconfigurable UWB

antenna, by adding two branchline directional couplers on a second FR4 sub-
strate layer introduced beneath the ground plane. Each coupler operates at the
center frequency of one of the channels.

Design of simple UWB patch at Channel 2

In order to achieve a patch antenna resonating along one UWB channel (500 MHz of
bandwidth), the approach described in [9] was employed, Figure 4.7 illustrates the
simulated structure for the operation in Channel 2 which is centered at 4 GHz.
The substrate used is FR4 with thickness t = 0.8 mm, the dimensions of the radiating

F IGURE 4.7: UWB antenna with capacitive feed and air gap.

element are L = W = 23 mm and were derived from standard patch antenna design.
The small feed patch has a width of wf = 1.6 mm and length of lf = 10 mm. It is
separated from the radiating patch by a gap distance g = 2 mm. The probe feed
center pin has a diameter of 1.3 mm and is placed at the center of the small feed
patch.
The effect of the air gap height (between the ground plane and the substrate) on
the bandwidth has been studied and illustrated in Fig. 4.8. It can be observed that
the bandwidth increases by increasing the air gap height "ta". A height of 8 mm is
optimal in this case as the reflection coefficient is under - 10 dB for all the frequency
range of Channel 2 (3.75 - 4.25 GHz).
74 Chapter 4. Localization with UWB antennas: the requirements of RTLSs

F IGURE 4.8: Air gap height effect on bandwidth of Channel 2

Design of UWB patch at Channel 5

The previous design is used and an L-shaped slot, of 2 mm width, is introduced to
the radiating element as illustrated in figure 4.9. The detached patch on the right
side of the slot is designed to resonate at the center frequency 6.49 GHz of channel 5
(6.25 - 6.75 GHz). It is a square patch with dimension W5 = 14 mm. A second small
patch (of length = 3 mm and width = 1.6 mm) and a probe are also introduced as a
feed to the CH5-patch.

F IGURE 4.9: Channel 2 configuration - Channel 5 configuration.

The same study of the bandwidth with varying air gap height was performed
and is illustrated in figure 4.10.

F IGURE 4.10: Air gap height effect on bandwidth of Channel 5.

Intuitively, the initial value was set to be of 8 mm as it was the optimal value
found for the CH2-patch previously. However, the resulting bandwidth does not
4.2. Frequency reconfigurable and circularly polarized patch antenna over dual
75
Ultra-wideband channels

cover the upper side of the channel 5 bandwidth, which should be until 6.75 GHz.
It can conversely be observed, that a lower value of the air gap, 6 mm, results in an
ideal bandwidth covering all of the channel 5 range.
These results show that, increasing the air gap height does increase the bandwidth
but its increase is limited at some point. This limitation depends on the desired
frequency of operation. It seems that the higher the desired frequency is, the lower
value of air gap height needed to achieve a certain bandwidth. A more extensive
and detailed parameter study of this concept has been described in [10].
Finally, since the CH2-patch and the CH5-patch are both in the same structure, the
air gap height needs to be the same for both cases, so a compromise of ta = 7 mm has
been chosen for the final reconfigurable structure.
It can also be noted that, other parameters, such as the small feed patch length "lf"
and distance between the radiating element and its feed patch "g" do not influence
greatly the bandwidth, but instead they have an effect on the impedance matching
as studied in [9] and [10]. As an example, the effect of the feed patch length on the
reflection coefficient can be observed in figure 4.11. From these results, the length of
the second feed patch is taken as lf-ch5 = 3 mm. The effect of the width of the L-slot
was also verified in Fig. 4.12, where it can be observed that this parameter does affect
the impedance matching, however its effect on the bandwidth is negligible [9].

F IGURE 4.11: Feed patch length effect on impedance matching for

Channel 5.

F IGURE 4.12: L-shaped slot width effect on impedance matching for

Channel 2.

Circular polarization
CP operation was achieved by :
76 Chapter 4. Localization with UWB antennas: the requirements of RTLSs

• Adding two other CP small patches on the top layer, each orthogonal to the
feed patch of the radiating element it corresponds to, as illustrated in Fig. 4.5.
The result of this, for each radiating element, is a feed patch on either the x-
axis (or y-axis) and a similar CP patch orthogonal to it placed on the y-axis (or
x-axis respectively). This allows to excite the two orthogonal modes TM01 and
TM10 (necessary for CP) in each of the radiating patches.

• Adding two branchline directional couplers, one operating at the 4 GHz fre-
quency and the other at 6.49 GHz, covering each the necessary 500 MHz chan-
nel bandwidth. To add both couplers, a second substrate layer was added
under the ground plane as illustrated in Fig. 4.4 and Fig. 4.6.

• The two feeding ports are represented by the two inputs of each coupler, as
seen in Fig. 4.4. While one output (from each coupler on the bottom layer)
was connected using copper pins, to one small feed-patch on the top layer, the
other outputs were connected to the small CP patches.

4.2.3 Results
The following are EM simulation results obtained using the final Anchor antenna
structure design of figures 4.4, 4.5 and 4.6 in CST Microwave Studio, with the two
frequency configurations.

UWB Frequency reconfigurablity

Figure 4.13 illustrates the simulated reflection coefficient of each coupler’s input, one
which feeds the CH-2 coupler and the other one which feeds the CH-5 coupler. These
results were obtained, when the copper connections (which model the p-i-n diodes)
were present or absent, for channel 2 and channel 5 respectively. The figure shows
good impedance matching for both configurations, as the reflection coefficients are
below - 10 dB on all of the channels’ bandwidths (3.75 - 4.25 GHz and 6.25 - 6.75
GHz).

F IGURE 4.13: Reflection coefficients of the reconfigurable antenna.

Circular polarization
CP in simulation was verified by simulation of the axial ratio (AR) value. Fig. 4.14
illustrates the AR on each channel’s frequency range for the main direction of radia-
tion. It can be seen that ultra wideband CP of the antenna on both configurations is
4.2. Frequency reconfigurable and circularly polarized patch antenna over dual
77
Ultra-wideband channels

verified, as the AR is under 3 dB for all channel 2 frequencies and under 3.37 dB for
all channel 5 frequencies.

F IGURE 4.14: Simulated axial ratio of the reconfigurable antenna

(main radiation direction).

4.2.4 Experimental characterization of the antenna

UWB Anchor antenna prototype
The UWB reconfigurable anchor antenna was fabricated in the lab using a 80.4 mm
x 46 mm x 0.8 mm FR4 substrate (for both layers) and 0.035 mm thick copper for
conductors and ground plane. The resulting antenna is illustrated in figures 4.15
and 4.16.

F IGURE 4.15: Top view of the UWB reconfigurable antenna proto-

type.

F IGURE 4.16: Bottom view of the UWB reconfigurable antenna pro-

totype.
78 Chapter 4. Localization with UWB antennas: the requirements of RTLSs

The final optimized dimensions of the antenna are : W2 = 23 mm, W5 = 14 mm,

air gap height ta = 7 mm, L-shaped slot width g = 2 mm, both feed and CP patches’
lengths lf2 = 10 mm and lf5 = 3 mm and widths w2 = w5 = 1.6 mm, interlayer pins
diameter = 1.3 mm.
In the following, simulation and measurement results will be compared.

Measured frequency reconfigurability

Figure 4.17 illustrates the simulated and measured reflection coefficients along chan-
nel 2 and channel 5. Despite a small frequency shift, good agreement between sim-
ulation and measurement results is observed, as the measured antenna impedance
is matched (S11 is inferior to - 10 dB) on both channels’ bandwidths. Thus, UWB
frequency reconfigurability of the antenna is verified.

F IGURE 4.17: Reflection coefficient at channel 2 and channel 5.

Measured circular polarization

The measured circular polarization was verified along channel 2 by two means :
measurement of the AR and measurements of the realized gain of both right-hand
and left-hand circular polarizations (RHCP and LCHP respectively). Figure 4.18
presents the obtained measured AR with respect to frequency. It can be observed
that the AR is below 4 dB along all channel 2 frequencies, which verifies the CP
operation.
4.3. Conclusion 79

F IGURE 4.18: Measured and simulated axial ratio of channel 2.

Figure 4.19 illustrates the realized gain of RHCP and LHCP. It can be seen that the
cross-polarization (here, LHCP) is suppressed at - 18 dBi for θ = 0° and at - 10 dBi for
a 3 dB beamwidth of almost 100°. While, it can be seen that the RHCP polarization
gain is of 5 dBi. These results verify the good performance of the UWB CP along
channel 2 frequencies and demonstrate that this CP has a right hand sense (RHCP).

F IGURE 4.19: Measured realized gain of RHCP and LHCP (at 4 GHz
and φ 90°)

4.3 Conclusion
This chapter presented a novel frequency reconfigurable UWB patch antenna with
circular polarization over two channels, one centered at 4 GHz and the other at 6.49
GHz. The approach of capacitive feeding mechanism and air-suspended substrate
were used to obtain a bandwidth of 500 MHz for each frequency configuration. Two
branchline couplers on a bottom layer were used to obtain wideband CP operation.
The antenna was fabricated and its reconfigurability was verified in terms of good
impedance matching on both channels’ frequency ranges. Its wideband CP oper-
ation along one of the channels (channel 2) was also verified, as it achieved 5 dBi
of realized gain for RHCP on the main directions of radiation. These performances
make this antenna a very suitable candidate for UWB localization systems’ Anchor
antennas as it operates in different channels, has a high gain compared to most exist-
ing UWB antennas, in addition to its circular polarization which is a very important
feature that helps tags and objects’ detection for small or large scale ranges.
80 BIBLIOGRAPHY

Bibliography
[1] Federal Communications Commission (FCC), First Order and Report, Revision of
the Commission’s Rules Regarding UWB Transmission Systems, Part. 15, 2002, pp.
02-48.

[2] C.C Cruz, J.R Costa, and C.A Fernandes, “Hybrid UHF/UWB antenna for pas-
sive indoor identification and localization systems,” IEEE Transl. Antennas and
Propagation, vol. 61, no. 1, pp. 354-361, January 2013.

[3] Y. Duroc, A.I. Najam, UWB antennas: design and modeling, IntechOpen, 2010.

[4] Decawave, DW1000 datasheet, ver. 1.6, 2016.

[5] S.S. Zhong, X.L. Liang,“Progress in ultra-wideband planar antennas,” J. of

Shanghai University. 11, pp. 95–101, 2007.

[6] J. Nan, J. Zhao, M. Gao, W. Yang, M. Wang and H. Xie, "A Compact 8-States
Frequency Reconfigurable UWB Antenna," in IEEE Access, vol. 9, pp. 144257-
144263, 2021, doi: 10.1109/ACCESS.2021.3122250.

[7] S. Tripathi, A. Mohan, and S. Yadav, “A compact frequency-reconfigurable frac-

tal UWB antenna using reconfigurable ground plane,” Microw Opt Technol
Lett. 59, pp. 1800–1808, 2017.

[8] M. Bitchikh, F. Ghanem, and S. Yadav, “A three-resolution UWB frequency re-

configurable antipodal vivaldi antenna for cognitive radios,” The 8th European
Conference on Antennas and Propagation (EuCAP 2014), pp. 3665-3668, 2014.

[9] G.M. Ridgers, J.W. Odendaal and J. Joubert, “Single-layer capacitive feed for
wideband probe-fed microstrip antenna elements," IEEE Transactions on An-
tennas and Propagation, vol. 51, no. 6, pp. 1405-1407, June 2003.

[10] M. Stanley, Y. Huang, H. Wang, H. Zhou, A. Alieldin and S. Joseph, “A Ca-

pacitive coupled patch antenna array with high gain and wide coverage for 5G
smartphone applications," IEEE Access, vol. 6, pp. 41942-41954, 2018.
 81

Chapter 5

Antenna Contributions to the

Improvement of Localization with
UWB Real-Time Locating Systems

5.1 Introduction
Indoor object localization and positioning is part of the space-awareness concept
which has seen a rising popularity in recent Internet of Things (IoT) research and
applications. This article presents a novel method to improve the localization per-
formance of ultra-wide band (UWB) real-time locating systems (RTLS) by improv-
ing the transmitting and receiving reader and tag antennas. Patch directional UWB
antennas with relatively higher gain compared to the generally used standard om-
nidirectional monopole UWB antennas have been exploited to achieve a larger lo-
calization range. Furthermore, the patch antennas were designed to have wide-
band circular polarization to achieve two objectives: a received power independent
of the orientation of the tagged objects that need to be detected, and the filtering
of unwanted multipath signals. A measurement campaign was conducted using a
commercially available RTLS with conventional antennas and then with the newly
designed antennas. A comparison between the localization results of the two an-
tenna types demonstrates an improved range with almost 100 m difference, received
power independent of tag orientation, and increased multipath mitigation with the
directional circularly polarized antennas.

5.2 State of the art

The future of real-time locating systems (RTLS) announces itself to be a prominent
topic as its applications find a niche in most domains such as industrial, scientific,
economic, and social [1, 2].
RTLS technology for high-accuracy localization is based on the impulse-radio
ultrawide bandwidth (IR-UWB) standard and was first introduced with the aim of
locating objects in indoor environments [3,4] such as inside buildings, mines [5], and
anywhere Global Positioning System (GPS) tends to fail owing to harsh propagating
conditions. For example, this is the case when the presence of obstacles and reflect-
ing objects leads to multipath components that affect the recognition of the desired
signal. Fortunately, the UWB standard is more robust against such propagation hin-
ders thanks to its wide spectrum (3.1 GHz - 10.6 GHz), where one UWB channel has
a bandwidth of 0.5 or 1 GHz [6–8].
Current industrial real-time locating architectures target the most accurate in-
door localization of objects [9]. Such architectures are primarly composed of one
 Chapter 5. Antenna Contributions to the Improvement of Localization with UWB
82
Real-Time Locating Systems

or several readers and tags. While attempting to locate a tagged object, they receive
continuously, real-time ranging signals in both the uplink (target -reader) and down-
link (reader -target), or only in the uplink, depending on the ranging method used
(two-way ranging (TWR) or time difference of arrival (TDoA), respectively [10]). Al-
though the highest accuracy is targeted during this continuous exchange, as long as
the locating system operates in a real imperfect environment, the system will most
likely fail to locate the object precisely at least once among its total ranging attempts,
especially in the case of changing environment due to the mobility of the tagged ob-
ject. For this reason, the goodness of an RTLS is not only defined by its accuracy,
but also by how consistently this accuracy is achieved. In some applications, such
as localization using radio frequency-identification (RFID) technology, meter level
accuracy is typically accepted as the objective is only to identify the tag and to know
it is present inside a defined area [11]. In contrast, centimeter level accuracy is re-
quired in other applications, such as intelligent ressource management, supply and
stock tracking in companies and personal medical monitoring in hospitals, which
call for the use of UWB technology.
In addition to consistent and high accuracy, research needs to further explore
other RTLS design specifications such as reading range, object orientation-independent
power, multipath mitigation ability, and detectability without prior knowledge of
the target direction.

F IGURE 5.1: Conventional real-time locating system configuration.

First, the reading range of the RTLS must be investigated. Because the popularity
of RTLS technology is typically for indoor use cases (from small surface retail centers
to large warehouses where logistics need to be performed), the need to extend the
reading range must be considered when choosing or designing the required RTLS.
If this issue is considered from a broad perspective, the range depends first on the
technology used for localization. If it is passive or semi-passive, such as RFID sys-
tems, then the achieved operating range will be limited to, typically, 10 m [12], as
it mostly consists of passive tags, powered through wireless power transfer [13, 14].
However, if the technology is UWB, which mostly consists of battery-powered de-
vices [15], the expected range can exceed 100 m, which is ten times greater than
the current passive locating devices’ range. If this issue is then considered from a
more specific and smaller perspective, that is, with an already chosen technology in
mind, the reading range depends deeply and mostly on the transmit power and link
5.2. State of the art 83

margin of the radio system. In this case, improving the transmit power fed to the
front-end antennas would require either increasing the transmitter power itself or
choosing an antenna with a higher gain compared to the state-of-the-art dipole-like
antennas commonly used in indoor locating systems.
The next step of this study consists of the design of RTLS antennas. Indeed,
these front-end components play a significant role in improving localization results,
not only in terms of reading range through gain adjustment but also in the ability
to enable object-orientation independent power [16] and multipath mitigation [17]
through polarization adjustment (from commonly linear to circular polarization). In-
deed, in [18], antenna polarization was taken into account to characterize the trans-
fer function of UWB transmitting and receiving antennas. For this reason, their opti-
mization should be considered, particularly when other devices and hardware in the
RTLS architecture cannot be optimized nor changed at all. Recently, for this purpose,
a novel circularly polarized ultrawide bandwidth high-gain antenna operating over
two 500 MHz channels was proposed by the same authors [19]. Moreover, the im-
portance of attenuating multipath components for UWB ranging applications was
highlighted in [20–24], thus, solutions such as the help of circular polarization in
reducing these components need to be investigated.
Thus, the main contributions of this chapter are :

• To demonstrate, through a real-scenario measurement campaign, the advan-

tages of using the realized directional circularly polarized UWB patch antenna
[19] in an available commercial RTLS system [25]. The measurements consist
in ranging with an UWB tag and reader in outdoor and indoor environments.

• To highlight the advantages of using a directional circularly polarized antenna

by the comparison of the achieved localization quality while employing the di-
rectional circularly polarized antenna, with its performance when using com-
mercial omnidirectional linearly polarized dipole-like antennas. This com-
parison between the two cases, is made in terms of the reading range, object
orientation-independent received power, and multipath mitigation.

• To investigate the effect of this antenna’s characteristics on the reading range.

For this purpose, an outdoor environment was privileged for the ranging mea-
surements, to benefit of unlimited distance for tests. The results showed that
the directional circularly polarized antennas achieve about 100 m of increased
range compared to the linearly polarized antennas.

• To investigate the effect of the antenna’s circular polarization on enabling the

independence of the received power from the relative orientations of the reader
and tag antennas. For this purpose, ranging was performed for different reader
and tag orientations, and an indoor environment was privileged to ensure the
validity of the results even in a complex environment, including walls and
objects. The received power was found to be independent of the orientations
of the reader and tag when using the circularly polarized antennas.

• To demonstrate the effect of the antenna’s circular polarization on the atten-

uation of the multipath signals during ranging measurements. The results
showed that, both in indoor and outdoor environments, the multipath is sig-
nificantly reduced in the case of circularly polarized antennas compared to
linearly polarized antennas.
 Chapter 5. Antenna Contributions to the Improvement of Localization with UWB
84
Real-Time Locating Systems

The remainder of this chapter is organized as follows. Section II describes the

UWB locating architecture flow. Section III analyzes the employed UWB antenna
design characteristics, followed by section IV in which the designed antenna results
and a comparison with the commercial antenna are presented. Section V addresses
the use case measurement campaign and scenarios, and discusses ranging measure-
ment results in detail through a comparison study. Finally, the conclusions are pre-
sented in Section VI.

5.3 Localization with RTLS technology

Real-time locating technology allows the inference of highly accurate position infor-
mation. It has emerged with the need to localize objects situated in harsh propaga-
tion environments, for example, trapped inside buildings or inside mines. Thus, it
is employed in indoor scenarios characterized by the presence of multipath signals,
reflections, and obstacles. To mitigate these problems and infer centimeter-level ac-
curacy of the localization, RTLS is typically based on UWB signals, as they allow the
precise measurement of the time-of-arrival (ToA) [9], from which the distance is then
calculated during the received signal post-processing phase.
Fig. 5.1 illustrates the RTLS configuration considered in this study. It is composed
of a reader-tag nodes couple that communicates through a two-way ranging (TWR)
method, which resorts to ToA measurements to allow the reader to infer the tag’s
location.

5.3.1 Ranging method

The TWR method allows for ranging, that is, distance estimation between the two
nodes. However, to infer tag position estimates of coordinates ( xt , yt ), at least three
readers performing ToA measurements are required. In contrast, there is an addi-
tional well-known method, that uses only one-way tag-to-reader communication to
obtain position information, which is TDoA method. This technique also requires at
least three readers and estimates the position of the tag by measuring the difference
in the time at which the signal from the tag reaches these readers [9, 10, 26].
In this study, the objective is to highlight the advantage of using more suit-
able antennas (180° directional radiation antennas with UWB circular polarization)
compared to commonly used UWB low-gain omnidirectional antennas with linear
polarization, to improve the performance of the general RTLS in terms of reading-
range, object orientation independent power and achieved multipath filtering. Con-
sequently, we focus on the TWR method to estimate the distance rt between the
reader and tag, as this objective needs to be initially verified for one reader-tag cou-
ple configuration to further seek its implementation in more complex architectures
that call for a higher number of fixed readers.

5.3.2 Instrumentation and configuration

In this section, we describe the waterfall of the UWB RTLS system. Table 5.1 below
lists the instrumentation components used during the ranging measurements.

Transmitter
In this study, the reader transmits a 500 MHz bandwidth UWB interrogation signal.
In the time domain, this signal is a pulse of frequency f c = 4 GHz with a pulsewidth
5.3. Localization with RTLS technology 85

TABLE 5.1: Adopted instrumentation

Reader UWB transceiver STMicroelectronics [25]

Tag UWB transceiver STMicroelectronics [25]
Antenna 1 Omnidirectional STMicroelectronics [25]
Antenna 2 180° radiation this work and [19]

of Tw = 3 ns generated by the electronics board. The pulse is repeated with a period

Tp and applied to an UWB antenna. According to STMicroelectronics documenta-
tion [25], the transmitted spectral power density measured at the antenna output has
a peak level of −36 dBm/MHz, which complies with the Federal Communications
Commission (FCC) mask limits [4, 6, 7]. This limit is imposed on the effective radi-
ated isotropic power (EIRP) and the antenna gain in the maximum direction of the
transmitting antenna is considered [4, 10].

Communication link
The tag antenna receives the interrogation signal, which enables the tag to transmit
its response signal after a fixed reply delay τd of the transceiver module. The TWR
method between the reader and tag is illustrated in Fig. 5.2, as explained in [9].

F IGURE 5.2: Two-way ranging principle.

Receiver
Because electromagnetic waves propagate in free space at a speed of light c = 3 ×
108 m/s, it is possible to determine the distance between the reader and tag. The
reader starts by having the instant t1 time which is known to the module, sends the
request packet to the tag and, then awaits the signal to return after a round trip time
(RTT) to obtain the time instant t4 . The response delay, τd is primarily known to the
device. Therefore, the time of flight between the reader and tag can be obtained from
the following equations [9]:
(t4 − t1 ) − τd
τf = (5.1)
2
RTT = t4 − t1 (5.2)
Using the obtained time of flight τ f , the distance can then be computed, and the
ranging is finalized.
 Chapter 5. Antenna Contributions to the Improvement of Localization with UWB
86
Real-Time Locating Systems

5.4 UWB Antenna design

The following section focuses on the antenna component and details the design mo-
tivation and steps.

5.4.1 Antenna design motivation

Typically, in the RTLS architecture based on UWB technology, as mentioned pre-
viously, both reader and tag antennas exhibit omnidirectional dipole-like radiation
behavior. Indeed, this is a key enabler for large coverage because both nodes can de-
tect each other regardless of their relative position. However, for most localization
techniques, and their applications inside buildings, retail wear-houses, or industrial
infrastructure, readers are mostly fixed on walls or any vertical surface of a device
such as a machine or a robot that goes around the facility to track items. This is the
case, for example, for RFID and RTLS solutions proposed by "Kathrein" or "Turck
Vilant" companies [27, 28].
Fixing a device with an omnidirectional antenna on a vertical surface would not
take advantage of this omnidirectionality because half of the radiated wave would
be absorbed on the back surface. Another case would be if an omnidirectional an-
tenna is placed on a metallic surface, which would not absorb the wave but would
disrupt the antenna performance particularly due to mismatch. Starting from this
point, the constraint of omnidirectional radiation could be lightened on antennas for
RTLS system readers, as only a 180° coverage would be sufficient to detect tags in the
surrounding environment. For this reason, a patch-type antenna, operating along
two 500 MHz UWB frequency channels (channel 2 and 5) was proposed in [19].
For details regarding UWB channels allocation in the frequency spectrum, Table 5.2
illustrates the lowband frequency UWB channels’ information, among which the
channels mentioned in this work are allocated.
TABLE 5.2: European UWB lowband frequency channels

Channel Center frequency (MHz) Bandwidth (MHz)

1 3494.4 499.2
2 3993.6 499.2
3 4492.8 499.2
4 3993.6 1331.2
5 6489.6 499.2

As observed, UWB channel 2 is centered at 4 GHz (3.75 GHz to 4.25 GHz), and
channel 5 is centered at 6.5 GHz (6.25 GHz to 6.75 GHz). Thus, if desired, the initial
antenna in [19] could be frequency reconfigured through p-i-n diodes to choose on
which channel to operate, giving it the capability to adapt to different commercial
RTLS UWB-based electronics modules provided for localization, which might not
be designed to work on the same UWB channel. Furthermore, this antenna has the
particularity of being circularly polarized along the UWB bandwidth of each channel
enabling object orientation-independent ranging.
To accommodate the commercial RTLS and UWB localization module used for
measurements (the B-UWB-MEK1 evaluation module [25] from former BeSpoon
company, acquired recently by STMicroelectronics), which operates by default on
channel 2, this channel was chosen as the channel along which the new antenna will
be designed to operate.
5.4. UWB Antenna design 87

Furthermore, because antenna frequency reconfigurability is not necessarily needed

here, a simplified one-channel version of the previous antenna was designed and
fabricated to operate along channel 2 and will be used on both reader and tag sides
in the measurement campaign demonstrated below in chapter. In the following, we
detail design method, characteristics, and results of the channel 2 antenna.

5.4.2 Antenna design

The objective of the design was a high-gain, UWB antenna with circular polarization
and 180° radiation coverage. The design steps are summarized as follows:

• To achieve relatively high gain and 180° coverage, a patch-type antenna was
chosen.

• Subsequently, a capacitive feed and coupling method was employed to achieve

UWB impedance matching over the requested channel.

• Finally, a directional coupler was integrated into the bottom substrate layer to
obtain circular polarization along the achieved antenna bandwidth.

The initial frequency reconfigurable antenna [19] profile is illustrated in Fig. 5.3 and
the new simplified version (one-channel), designed here, is shown in Fig. 5.4. The
adopted capacitive feed technique [29] is aimed at probe-fed patch antennas, and
consists of exciting a small element placed on the same substrate layer as the radiat-
ing element and at a gap distance from it. The dielectric substrate was suspended in
air as shown in Fig 5.3.

F IGURE 5.3: Side view of the frequency reconfigurable UWB antenna

[19].

F IGURE 5.4: Simplified circular polarized antenna version, operating

on one UWB channel, with radiating element (top layer) and direc-
tional coupler (bottom layer).
 Chapter 5. Antenna Contributions to the Improvement of Localization with UWB
88
Real-Time Locating Systems

As illustrated in Fig. 5.4, the new antenna is composed of only one radiating
element and one coupler both designed at the center frequency f = 4 GHz. The
dimensions of the radiating element are W = L = 23 mm and were derived from the
standard patch antenna design. The two capacitive feed lines ensure two orthogonal
feeds and thus circular polarization. They are both of width w f = 1.6 mm and
placed at a gap distance from the radiating element d = 2 mm. These two lines are
both fed by vias attached to the coupler outputs which are situated at the bottom
substrate. The vias start from this substrate and go through the ground plane, the
air gap (ensuring bandwidth), and finally to the upper substrate.

5.5 Antenna simulation and experimental results

This section presents the performance of the antenna achieved by the simulation and
measurements in terms of :

• Reflection coefficient (impedance matching) along the frequencies of UWB chan-

nel 2.

• Directivity and gain.

• Radiation pattern.

• UWB circular polarization along 180° coverage.

These results will allow the implementation of this antenna in the RTLS architecture
for performing ranging measurements.
The circularly polarized UWB antenna operating in channel 2, was fabricated using
FR4 substrate of height 0.8 mm for both substrate layers. The antenna prototype is
shown in Fig. 5.5. The air gap betweeen the ground plane and the upper substrate
was replaced by foam whose dielectric constant is the very close to that of air. The
structure has two input ports, which are the inputs of the coupler, one of which is
isolated during measurements using a 50 Ω match load.

F IGURE 5.5: Prototype of the circularly polarized antenna operating

over UWB channel 2.

5.5.1 UWB impedance matching

Fig. 5.6 shows the result of the simulated and measured reflection coefficient of one
of the antenna ports (S11) with respect to the frequency, with the chosen air gap
height h air = 7 mm, considering the symmetry of the structure, the results on port 2
(S22) are the same in both measurement and simulation. With reference to −10 dB,
the antenna impedance is matched along a bandwidth covering all UWB channel
2 frequencies. Furthermore, it can be observed that the antenna bandwidth covers
more frequencies than the desired channel bandwidth, as it starts around 2.5 GHz
5.5. Antenna simulation and experimental results 89

F IGURE 5.6: Antenna reflection coefficient.

and extends until around 4.8 GHz, this is a result of the chosen air gap height of
7 mm as this value is not the minimum limit value to achieve the 500 MHz range.
This choice was made to cover the maximum bandwidth possible for other future
use of the antenna and was not related to any other restrictions.

5.5.2 Directivity and gain

The antenna was measured in an anechoic chamber and Fig. 5.7 shows the result-
ing measured antenna directivity and gain with respect to frequency. The observed
gain is of approximately 6.5 dBi and is stable along all channel 2 frequencies (3.75 to
4.25 GHz). The directivity follows the same curve but is naturally higher than the
gain, maintaining a value of 8 dBi along the channel frequencies.
It must be pointed out that maintaining a stable gain along all the desired chan-

F IGURE 5.7: Measured antenna directivity and gain over the UWB
channel 2 frequencies.

nel frequencies is important for UWB antennas to exhibit the same behavior while
operating with UWB electronic devices.

5.5.3 Radiation pattern and circular polarization

The measured radiation pattern is presented in terms of polarization gain in the
azimuth plane and plotted in Fig 5.8 and Fig 5.9. This radiation is measured at the
center frequency f = 4 GHz varying the azimuth angle, for two elevation angles
 Chapter 5. Antenna Contributions to the Improvement of Localization with UWB
90
Real-Time Locating Systems

0° and 90°. As shown in Fig. 5.8, the circular polarization of the antenna is right-
handed (RHCP) because this polarization component has a gain of 6 dBi along the
main direction of the radiation (around azimuth 0), which represents a difference of
+ 16 dB with the contrapolar component in the same direction. Indeed, the cross-
polarization, that is, the left-handed (LHCP) polarization, is at least −10 dB lower
for any azimuth direction between -50° and 50°.

F IGURE 5.8: Measured azimuth antenna radiation pattern (polarisa-

tion gain) at center frequency 4 GHz, for elevation angles 0° and 90°.

5.5.4 Comparison with commercial antenna

The purpose of this work is to design a more suitable antenna optimized for real-
time localization UWB commercial devices. A comparison with a commercial an-
tenna belonging to the UWB evaluation kit from BeSpoon company (recently STMi-
croelectronics) [30] is reported in this section.
The radiation patterns of both the designed and commercial antennas are shown
in Fig 5.9 and Fig 5.10 respectively. First, we observed the radiation behavior of the
designed antenna through the radiation of its dominant RHCP polarization. It can
be observed that the antenna radiates directionally with 180° coverage. It specifically
maintains a high gain of 6 dBi along 70° azimuth centered at 0° for both elevation
planes (0° and 90°, as shown in Fig. 5.9.a and 5.9.b respectively). Back lobes are
attenuated on the back surface of the antenna to less than −8 dB.
As shown in Fig. 5.10, as the commercial antenna is linearly polarized (vertical
polarization), it has E-plane and H-plane radiations. We intuitively associate its E-
plane to the elevation plane and its H-plane to the azimuth plane. As the designed
antenna radiation was traced with varying azimuth, we compare Fig. 5.9 RHCP ra-
diation to the commercial antenna H-plane in Fig. 5.10 (varying azimuth, shown on
the right figure). It can be observed that the commercial antenna for channel 2 has
omnidirectional radiation along 360° with a gain of 2 dBi. Conversely, for the same
gain, the designed antenna exhibits directional radiation along 95° azimuth. How-
ever, it has a higher gain along 70° in the azimuth plane, which includes the main
directions of radiation when the antenna is fixed to a wall or vertical surface. The
main differences between the antennas are presented in Table 5.3.
5.5. Antenna simulation and experimental results 91

F IGURE 5.9: Designed circularly polarized antenna’s measured radia-

tion pattern at frequency 4 GHz in the two polarizations. (a) Elevation
0° (b) Elevation 90°.

F IGURE 5.10: Commercial linearly polarized antenna’s measured ra-

diation pattern at channel 1, 2 and 3 [30].

TABLE 5.3: Antenna radiation comparison

Antenna 1 (commercial) Antenna 2 (this work)

UWB channel 2 UWB channel 2
Coverage (360° azimuth) Directional cov. (70° azimuth @ 5 dBi)
Low gain (2 dBi) High gain (6 dBi @ 60° azimuth)
Linearly polarized Circularly polarized
 Chapter 5. Antenna Contributions to the Improvement of Localization with UWB
92
Real-Time Locating Systems

TABLE 5.4: Antennae characteristics

Charact. / Antenna Antenna 1 (commercial Antenna 2 (this work and

[25, 30]) [19])

Frequency range 2 − 4.77 GHz 3.75 − 4.25 GHz

Reflection coeff. < −8 dB < −10 dB

Gain 2 dBi 6 dBi

Radiation Omnidirectional 180° directional

Polarization type Linear (vertical) UWB Circular (Right-

hand)

5.6 Measurement campaign and results

The experimental setup and measurements are presented in detail in this section.
The objective is to demonstrate the improved efficiency of an RTLS using TWR
method, to locate a tagged object by adjusting the antenna specifications to the local-
ization method’s needs. A measurement campaign was conducted using the STMi-
croelectronics UWB transceiver as an anchor and tag, and TWR between the two
was performed in two cases:

• First, employing typical commercial UWB omnidirectional monopole antennas

with linear polarization.

• Then, employing UWB directional patch antennas with circular polarization.

Measurements were repeated with the two antenna types for two different environ-
ments, outdoor and indoor, depending on the studied parameter (reading range,
object orientation independence, and multipath mitigation). Finally, both antenna
cases are compared and improvements are discussed.
The different characteristics of the antennas employed in these measurements
are reported in Table 5.4 and, more specifically, in Table 5.3.

5.6.1 Outdoor scenario for reading-range

Although RTLS localization specializes in locating objects indoors because of its ro-
bustness to harsh propagating environments, this part of the measurements was
realized in an outdoor environment because of space constraints inside buildings,
in order to estimate the maximum reading range of the RTLS at hand, for the two
employed antenna types. For this measurement, the transceiver tag was placed in a
fixed position, and the transceiver reader was moved. The approach is to start with
the reader as close as possible to the tag and then move away until the reader can
no longer detect the tag. Thus, it is no longer outputting the ranging results. These
ranging results appear on the computer to which the reader is attached as data ac-
quisition is possible through UART on USB, which is also used to power the reader.
The tag is powered by a power bank unit. The obtained results of ranging are as
follows:
5.6. Measurement campaign and results 93

• The time between measurements in milliseconds, knowing that for the TWR
mode, the transceiver is able to make a number of measurements up to 204/s.

• Device ID of the reader board itself.

• Device ID of the tag board.

• The link quality indicator (LQI), which is the ratio of the measured received
signal strength to the already known saturation signal strength of the transceiver,
it is computed as a percentage.

• The corresponding reader-tag distance in meters.

Figure 5.11 illustrates the measured LQI with respect to the measured distance
while moving further away from the tag.

F IGURE 5.11: Comparison of the RTLS reading range when using the
commercial antenna and the designed antenna.

Although the objective is to evaluate the performance range-wise, for clarity, the
reading range is plotted on the horizontal axis. Indeed, it is easier to observe that the
measured maximum reading distance is 82 m when using the commercial linearly
polarized (LP) omnidirectional antennas, whereas it is 175 m when using this work’s
directional CP patch antennas. Thus, an important difference of almost 100 m in the
maximum reading range was obtained owing to the antenna gain improvement.
It is important to note that, even in line-of-sight (LOS) conditions, the environment
and the initial position of the fixed tag or the position of the mobile reader during
the campaign leads to variations in the measurement results, such as the maximum
range obtained, due to the random nature of the radio propagation channel. This
is to highlight that, for example, the maximum range obtained with the commercial
transceiver and antennas here of 82 m is indeed different from the datasheet, which
claim a maximum range of 600 m, as typically this value is obtained either theo-
retically or through LOS measurements within a better propagating environment to
obtain the best result possible. In contrast, the measurements in this study were per-
formed in a completely realistic and imperfect outdoor environment. Indeed, other
measurement campaigns that we performed with commercial transceivers and an-
tennas were able to exceed 85 m and achieved a maximum range of approximately
120 m in a real environment, which is still significantly lower than the 175 m achieved
by the designed antenna. In conclusion, to compare the performance of RTLS pa-
rameters such as range, it is essential to ensure that the experimental setup and
 Chapter 5. Antenna Contributions to the Improvement of Localization with UWB
94
Real-Time Locating Systems

environmental characteristics are as closely matched as possible to obtain accurate

comparable results.

5.6.2 Indoor scenario for object orientation independence

The objective of the following experiment is to detail and verify the robustness
brought to the RTLS by the circular polarization characteristics of the designed an-
tennas. This robustness is mainly against two hindering factors in radio propaga-
tion, namely, the multipath signals due to reflections against surfaces or objects, and
the received signal attenuation due to polarization mismatch between reader and
tag (which depends on tag-reader relative orientation).
The reason why an indoor scenario was preferred for this experiment is that the
RTLS has to prove itself robust in a harsh environment (presence of objects and ob-
stacles), as this would mean it will also have good performance in outdoor scenarios
that tend to have fewer constraints, such as the absence of walls.
The considered indoor environment was the corridor of the LEAT laboratory that is
50 m long. The tag was mobile and the reader was fixed at one corridor end. Steps
of 15 m were set using a laser telemeter to determine the true absolute distances at
which the ranging measurements were taken. The reader antenna is kept on a "ver-
tical" configuration and the tag antenna orientation is first set to vertical (0°), and
then changed to horizontal (90°) orientation.
Fig. 5.12 and 5.13 illustrate the orientation configurations of the two antennas, the
commercial linearly polarized antennas and designed circularly polarized patch an-
tennas, respectively.

F IGURE 5.12: Two-way ranging with the commercial antenna, char-

acterized by 360° omnidirectional radiation and linear polarization.

F IGURE 5.13: Two-way ranging with the designed antenna, charac-

terized by 180° directional radiation and circular polarization.

To measure the polarization-mismatch independence of the received power be-

tween the reader and tag, a comparison between the LQI results obtained by ranging
5.6. Measurement campaign and results 95

(that is, distance measurement) with reader-tag antenna co-orientation and cross-
orientation is required. The following steps are followed for each distance point:

• Measurement of the LQI (by ranging) with reader and tag antennas having the
same orientation dcop (for example, vertical reader - vertical tag).

• Measurement of the LQI with reader and tag antennas with orthogonal orien-
tations dcrossp (for example, vertical reader - horizontal tag).

F IGURE 5.14: Object orientation-independence of received power,

comparison between linearly and circulary polarized antennas.
Reader antenna always in vertical orientation.

Fig. 5.14 illustrates the LQI as a function of the measured distance for the dif-
ferent antenna cases. LQI is proportional to the received power. These results were
obtained by ranging indoors to up to 35 m, with the reader antenna always fixed in
a vertical orientation.
By comparing the LQI achieved in the figure, it can be observed that, in the case
of the circularly polarized reader and tag antennas, the LQI (and thus the received
power) is almost identical at all distances, regardless of the orientation of the tag
antenna, that is, regardless of wether it’s the same as the orientation of the reader
antenna.
Differently, in the case of linearly polarized antennas used in both the reader and tag,
the received power is very different between co-polarization and cross-polarization
cases (blue and orange curves), along all distances. Here, the difference between
their LQIs varies between 20 % and 50 %. This is a significant amount of power
attenuation caused by polarization-mismatch, which greatly deteriorates the RTLS
performance, as some objects may not be detected if their antennas is not oriented
in the same orientation as the reader antenna.
It is important to note that polarization-mismatch induced power attenuation (which
can occur if the antennas are linearly polarized and have cross-orientations), for
longer distances, especially closer to reader sensitivity, can also negatively affect the
reading range.
It can also be observed in the same figure that the LQI achieved by the circularly
polarized antennas is always ≈ 20 % higher than the LQI obtained in the best case
 Chapter 5. Antenna Contributions to the Improvement of Localization with UWB
96
Real-Time Locating Systems

scenario of linearly polarized antennas (ie. both the reader and tag antennas verti-
cal). This difference is mainly owing to the higher gain of the designed directional
antennas.

5.6.3 Multipath mitigation with circular polarization

Fig. 5.15 illustrates further indoor and outdoor ranging measurements, where the
objective is to investigate wether circular polarization helps mitigate multipath re-
flection signals, which tend to be more prevalent in indoor environments because of
the presence of walls and usually more objects. Behind the reasoning that circular
polarization reduces the number of multipaths, is that if the transmitting and receiv-
ing antennas both have same-sense circular polarization (here, right-hand sense),
the receiver antenna filters out reflected signals from metallic surfaces as they are re-
flected with circular polarization of opposite-sense [17]. Furthermore, multipath is
undesired in localization applications because of the possibility of path-overlap [31],
which consists of reflection signals overlapping with the first direct-path signal, from
which the time-of-arrival is computed. This path-overlap would then cause non-
recognition of the direct path signal which will lead to errors in the time-of-arrival
measurement and thus errors in distance estimation.
Thus, we evaluate the LQI curves, obtained through ranging, to determine the mul-
tipath effect on the ranging measurement and compare it between the two antenna
cases, with linear and circular polarizations.

F IGURE 5.15: Multipath and antenna gain effects on received signal

strength in ranging.

From the results shown in Fig. 5.15, the curves of the linearly and circularly polar-
ized antennas follow almost the same high-and-low patterns at the same distances,
in both indoor and outdoor measurements. This is because the same multipath sig-
nal components were present at these distances (that is, identical environment at
each distance). Thus, this multipath will affect the propagation in the same manner
for both antenna types, indoors and outdoors.
Furthermore, if we compare the linearly polarized antenna curve with the circularly
polarized antenna curve (both indoors and outdoors), it can be observed that the
LQI and thus received power dips and peaks are less significant and have a low level
of variation when using circular polarization compared to linear polarization. This
5.7. Conclusion 97

observation demonstrates that circular polarization with the same direction (here,
Right Hand) in both reader and tag antennas helps to filter reflections, precisely
those that arrive with opposite circular polarization (here, Left Hand). This can be
observed for both the indoor and outdoor cases.

5.7 Conclusion
This chapter investigated the positive effects and advantages of using directional
circularly polarized UWB antennas to improve RTLSs in terms of reading range,
object-orientation independent received power, and multipath mitigation. Real-
environment measurements were conducted in indoor and outdoor scenarios using
a commercially available RTLS. The results demonstrate that the reading range was
improved in the case of the designed directional antennas compared to conventional
omnidirectional UWB antennas by almost 100 m owing to the higher gain, as well as
to the suppression of the cross-polarization mismatch effect on the received signal
strength by circular polarization. Object-orientation independent received power
was also observed in the case of circularly polarized antennas, especially in harsh in-
door environment. In contrast, power attenuation caused by polarization mismatch
was observed in the case of linearly polarized antennas. Furthermore, circularly
polarized antennas were observed to help multipath mitigation in both indoor and
outdoor scenarios, as the LQI curves demonstrated less harsh variations compared
to linearly polarized antennas.
Thus, we conclude that locating systems can be significantly improved through
directionality and circular polarization of antennas compared with the currently
widely used omnidirectional linearly polarized UWB monopole antennas. The type
of antennas presented in this work can not only improve time-based locating sys-
tems, such as the system employed here, but also received signal strength indicator
(RSSI) based locating technologies, such as RFID locating, as it would help increase
the power and thus help additionally provide precise localization information.
98 BIBLIOGRAPHY

Bibliography
[1] L. Yang and G. B. Giannakis, "Ultra-wideband communications: an idea whose
time has come," in IEEE Signal Processing Magazine, vol. 21, no. 6, pp. 26-54,
Nov. 2004.
[2] G. R. Aiello and G. D. Rogerson, "Ultra-wideband wireless systems," in IEEE
Microwave Magazine, vol. 4, no. 2, pp. 36-47, June 2003.
[3] R. A. Scholtz, “Impulse Radio: How it works,” IEEE Communications Letters,
Vol. 2, NO. 1, Jan. 1998.
[4] FCC, “Revision of part 15 of the commission’s rules regarding Ultra-Wideband
transmission systems,” Federal Communications Commission, Washington,
D.C., Tech. Ref. ET Docket 98-153, Apr. 2002.
[5] Y. Rissa, L. Talbi, and M. Ghaddar, “Experimental characterization of an UWB
propagation channel in underground mines,” IEEE Trans. Antennas Propag.,
vol. 60, no. 1, pp. 240–246, Jan. 2012.
[6] "IEEE Standard for Low-Rate Wireless Networks," in IEEE Std 802.15.4-2020
(Revision of IEEE Std 802.15.4-2015) , vol., no., pp.1-800, 23 July 2020.
[7] "IEEE Standard for Low-Rate Wireless Networks–Amendment 1: Enhanced Ul-
tra Wideband (UWB) Physical Layers (PHYs) and Associated Ranging Tech-
niques," in IEEE Std 802.15.4z-2020 (Amendment to IEEE Std 802.15.4-2020),
pp.1-174, 25 Aug. 2020.
[8] M. Ghavami, “Ultra Wideband Signals and Systems in Communication Engi-
neering”, 1st Edition, Wiley, 2004.
[9] S. Frattasi, F.D. Rosa, and D. Dardari, Ultra-wideband Positioning and Tracking. In
Mobile Positioning and Tracking, ed. S. Frattasi and F.D. Rosa, 2017.
[10] N. Decarli and D. Dardari, "Time Domain Measurements of Signals Backscat-
tered by Wideband RFID Tags," in IEEE Transactions on Instrumentation and
Measurement, vol. 67, no. 11, pp. 2548-2560, Nov. 2018.
[11] T. Sanpechuda and L. Kovavisaruch, "A review of RFID localization: Applica-
tions and techniques," 2008 5th International Conference on Electrical Engineer-
ing/Electronics, Computer, Telecommunications and Information Technology,
pp. 769-772, 2008.
[12] K. Finkenzeller, "RFID Handbook: Fundamentals and Applications in Contact-
less Smart Cards and Identification," 3rd ed. Wiley, 2010.
[13] D. Dardari et al., "An Ultra-Low Power Ultra-Wide Bandwidth Positioning Sys-
tem," in IEEE Journal of Radio Frequency Identification, vol. 4, no. 4, pp. 353-
364, Dec. 2020.
[14] A. Sidibe, M. A. Kassem, A. Takacs, J. Mennekens and J. Dachy, "Autonomous
and Battery-Free Wireless Tag combining UWB and BLE Technology," 2022
Wireless Power Week (WPW), Bordeaux, France, 2022, pp. 207-212.
[15] D. Dardari, A. Conti, U. Ferner, A. Giorgetti and M. Z. Win, "Ranging With
Ultrawide Bandwidth Signals in Multipath Environments," in Proceedings of the
IEEE, vol. 97, no. 2, pp. 404-426, Feb. 2009.
BIBLIOGRAPHY 99

[16] C. Su, Y. Liu, L. Liu, M. Yang, H. Zhao, and X. Yin, “Experimental Evaluation
of Multipath Mitigation in TDOA-Based Indoor Passive Localization System
Using A Beam Steering Broadband Circular Polarization Antenna,” Electronics,
vol. 7, no. 12, p. 362, Dec. 2018,
[17] R. Szumny, K. Kurek and J. Modelski, "Attenuation of multipath components
using directional antennas and circular polarization for indoor wireless posi-
tioning systems," European Radar Conference, 2007, pp. 401-404.
[18] X. Qing, Z. N. Chen, and M. Y. W. Chia, “Characterization of ultrawideband
antennas using transfer functions,” Radio Sci., vol. 41, no. 1, pp. 1–10, Feb. 2006.
[19] A. Benouakta, F. Ferrero, L. Lizzi and R. Staraj, “Frequency Reconfigurable and
Circularly Polarized Patch Antenna Over Dual Ultra-wideband Channels,” in
16th European Conf. on Antennas and Propagation (EuCAP), Madrid, Spain, 2022,
pp. 1-5.
[20] A. F. Molisch, D. Cassioli, C.-C. Chong, S. Emami, A. Fort, B. Kannan, J.
Karedal, H. G. Schantz, K. Siwiak, and M. Z. Win, “A comprehensive standard-
ized model for ultrawideband propagation channels,” IEEE Trans. Antennas
Propag., vol. 54, no. 11, pp. 3151–3166, Nov. 2006.
[21] A. F. Molisch, “Ultra-wide-band propagation channels,” Proc. IEEE, vol. 97, no.
2, pp. 353–371, Feb. 2009.
[22] A. Fathy, A. Yahya, and H. Ragai, “Deterministic UWB channel modeling using
ray tracing approach,” in Proc. Int. Conf. Eng. Technol. (ICET), Cairo, Egypt,
Apr. 2014, pp. 1–4.
[23] P. Pagani, F. T. Talom, P. Pajusco, and B. Uguen, "Ultra Wide Band Radio Prop-
agation Channels: A Practical Approach,". Hoboken, NJ, USA: Wiley, 2008.
[24] M. Schmieder, F. Undi, M. Peter, E. Koenig, and W. Keusgen, “Directional wide-
band channel measurements at 28 GHz in an industrial environment,” arXiv
preprint arXiv:1907.01346, 2019.
[25] Evaluation kit for the B-UWB-MOD1 ultra-wideband module. [Online]. Available :
https://www.st.com/en/wireless-connectivity/b-uwb-mek1.html
[26] D. Dardari, N. Decarli, A. Guerra, A. Al-Rimawi, V. M. Puchades, G. Prati, M.
De Angelis, F. Fraboni, L. Pietrantoni, "High-Accuracy Tracking Using Ultraw-
ideband Signals for Enhanced Safety of Cyclists," in Mobile Information Systems,
vol. 2017, Article ID 8149348, 13 pages, 2017.
[27] Katherein solutions. [Online]. Available : https://www.kathrein-solutions.com/
[28] Turck vilant. [Online]. Available : https://turckvilant.com/
[29] G.M. Ridgers, J.W. Odendaal and J. Joubert, “Single-layer capacitive feed for
wideband probe-fed microstrip antenna elements," IEEE Transactions on An-
tennas and Propagation, vol. 51, no. 6, pp. 1405-1407, June 2003.
[30] BeSpoon UWB Omnidirectional antenna BSP-UWBA-Om4G, datasheet V.1.0. [On-
line].
[31] Y. Shen and M. Z. Win, "Effect of Path-Overlap on Localization Accuracy in
Dense Multipath Environments," 2008 IEEE International Conference on Com-
munications, Beijing, China, 2008, pp. 4197-4202.
 101

Chapter 6

Multi-Standard UWB-LoRa
High-Accuracy Long-Range
Localization

6.1 Introduction
Long-Range Wide-Area Networks (LoRaWAN) allow the transmission of data via
radio link from sensors, which are potentially isolated or difficult to access, to gate-
ways and servers that are connected to cellular networks for data processing, ex-
change, or relay, with low transmission power. This concept employs Long-Range
(LoRa) modulation and has led to the emergence of many applications for the moni-
toring and tracking of objects. However, due to its characteristic of a low data rate for
low-power communication, the transmission of information with LoRa tech-nology
is not suitable for the fast real-time monitoring of data. Additionally, due to its
narrow bandwidth, an attempt to perform localization through the LoRa modula-
tion technique will result in very limited accuracy because of its inability to resolve
multipath problems. Thus, in this chapter, we propose a multi-standard Ultra-Wide
Bandwidth (UWB) and LoRa end-device that is capable of measuring location with
high accuracy using UWB technology and then transmitting the location informa-
tion through LoRa method to gateways and the Internet of Things Network. The
results of measurements in indoor and outdoor scenarios show a UWB localization
accuracy that is of sub-meter level, being between 10 and 33 cm, and a UWB range
of 124 m in Line-of-Sight (LOS) and 55 m in Non-Line-of-Sight (NLOS) applications,
respectively.
102Chapter 6. Multi-Standard UWB-LoRa High-Accuracy Long-Range Localization

The connectivity of objects to gateways using the LoRaWAN protocol has al-
lowed the emergence of many monitoring applications, such as intelligent resource
management, predictive maintenance, supply and stock tracking, object and ani-
mal tracking and personal medical monitoring [1]. This suggests infinite monitoring
possibilities depending on the physical variable detected via the sensor. Besides
these monitored variables, it is often necessary for gateways to monitor the loca-
tion of the sensor itself [2], especially if it is integrated into a mobile object or tag,
for example, if this latter feature is moving inside an indoor environment. The task
of localizing a mobile sensor could be challenging for the LoRa working scheme,
as this type of communication was not orig-inally optimized for such an applica-
tion [3], especially if the localization is required to be of high accuracy or expected
to accommodate both indoor and outdoor scenarios. In contrast, the main advan-
tage of LoRa communication in LoRaWAN networks is that the sensing information
can be transmitted along long ranges in the order of kilometers, typically ranging
from 5 km in urban areas up to 15 km in remote areas, from the sensor to the gate-
ways [4, 5]. Indeed, the LoRa technique is used to ensure deep-in communi-cation
among a large number of devices that have low power requirements and collect and
transmit small amounts of data [5,6]. Furthermore, LoRaWAN networks have a high
capacity and can handle millions of messages from thousands of gateways. How-
ever, despite these advantages, this technology is still not suitable for gateways to
locate mobile sensors precisely and as often as is necessary [7]. Indeed, to ensure
the low consumption of power, LoRa sensors send data with low packet rate [8, 9],
that is, typically one or two packets are sent per day. This is not suitable if the object
monitored is mobile in its environment and requires real-time monitoring, or at least
partial real-time monitoring, which cannot be achieved with the relatively low data
rate of the LoRa technique. Furthermore, if the monitored object needs to be located
with high accuracy and/or if it were in an indoor environment, we would require
radio communications with wide frequency band-widths to resolve multipath prob-
lems [10]; however, this is not the case for narrowband LoRa communications, which
have bandwidths of a few MHz.
Conversely, Ultra-Wide Bandwidth (UWB) is the pervasive technology nowadays
when it comes to locating objects or tracking assets or any type of targets, espe-
cially those in complex indoor environments, such as inside buildings, industrial
infrastruc-tures, hospitals, airports, construction sites [11, 12], etc. It is effectively
used in time-based ranging and localization techniques such as one-way or two-way
ranging with Time-of-Flight (ToF) and Time-Difference of Arrival (TDoA) [13]. It al-
lows high-accuracy distance and position estimates, notably due to the transmission
of the ranging infor-mation along a large frequency bandwidth of 500 MHz or more
between a reader and a target. Its constraint is that it is considered a short-range
communication technology despite having reading ranges of typically 100 to 200 m
in Line-of-Sight applications [14], which are mostly enough for indoor use cases but
cannot adapt in networks deployed in highly remote outdoor areas, where objects
can be spaced with more than those distances, i.e., such as in LoRaWAN networks.
Thus, in this work, we propose combining both LoRa and UWB technologies as
complementary Internet of Things (IoT) schemes into one transceiver board to ex-
ploit both of their principal features. This combination enables LoRa gateways to
locate the sensors that belong to it if they are mobile in their environment, with the
high-accuracy and real-time availability of their position information. It also allows
for UWB targets to be located at long-range data links such as those of LoRa links.
6.2. Related work 103

This solution consists of employing UWB technology in the mobile target LoRa sen-
sor to allow LoRaWAN gateways to track it in real time with high accuracy. The re-
sult is a multi-standard UWB-LoRa transceiver, which can work as a sensor-tag or a
reader depending on where it is placed in the communication chain. For this reason,
we propose equipping the mobile LoRa sensor with UWB technology (LoRa-UWB
sensor-tag) and placing the proposed UWB-LoRa reader device as an intermediate
node between the target object and the gateway. The reader receives the location
information from the sensor-tag in real time through UWB ranging with ToF and
sends it to the gateway through LoRa signals.
The remainder of this chapter is organized as follows: Section 2 describes the related
work in the context of the localization with LoRa technology Section 3 first describes
the overall UWB-LoRa localization system before describing the design and struc-
ture of the proposed UWB-LoRa transceiver, and the last subsection focuses on the
design of the antenna structure that is included in the transceiver. Section 4 presents
the characteri-zation results of both antennas from their reflection coefficient to their
radiated patterns, followed by Section 5, in which the achieved localization of the
transceiver is discussed and characterized by the UWB localization range and ac-
curacy between the sensor-tag and the reader, as well as the location information
transmission from the reader to the network. Finally, Section 6 concludes on the
paper contributions and results.

6.2 Related work

Previous research has realized localization attempts via LoRa technology, mostly us-
ing the Time Difference of Arrival (TDoA) method [7, 15, 16], which requires at least
three gateways to infer the location of the object. For example, in [16], a number of
sets of messages sent via the target sensor were used to calculate the TDoAs of these
messages and perform location estimation at the gateway level, the authors assessed
the efficiency of the performance of the localization and the results indicated that the
localization error was greatly affected by the noise of the received timestamps at the
base stations. In [17], a similar procedure was followed to implement a LoRaWAN
tracking system that was capable of exploiting transmitted packages to calculate the
current position, and using LoRa signals and applying a multilateration algorithm
on the timestamps received at the gateways, the results demonstrated that it can be
feasible to locate a device in a static spot with an accuracy of around 100 m, though
this accuracy is not good enough for indoor scenarios.
The reason that the times of arrival were not precise enough is the combination of
undesired multipath signals at the receiving gateways, with these signals being due
to the environment and unavoidable, especially because LoRa signals’ widths in
time (142 ns for the 863–870 MHz channel) are not narrow enough for the multipath
to be dis-tinguished from the desired path signal. In addition to problems related
to low accuracy, the previously cited studies all employed three or more gateways
to infer the positions of the sensors via the TDoA method, which represents another
constraint, as the already deployed LoRaWAN networks do not always geographi-
cally provide this condition.
Other studies employed the ToF method to measure the distance between one gate-
way and the object [2]; however, the server can only calculate this distance based on
the packet metadata provided by the gateway, and these metadata are only sent with
low data rates, which make real-time monitoring difficult to achieve. Furthermore,
in [18], researchers proposed a ToF-based localization method using a fingerprint
104Chapter 6. Multi-Standard UWB-LoRa High-Accuracy Long-Range Localization

map to handle the accuracy issues and reduce the localization error caused by the
noise and multipath; however, the fingerprinting method depends on the environ-
ment and needs to be updated for the location estimation to work accordingly, which
makes it effort and time consuming.
Other methods include the using Received Signal Strength Indicator (RSSI) to in-
fer the distance. For example, in [19], researchers measured the RSSI in an indoor
envi-ronment with a short distance under both LOS and NLOS conditions, and the
results showed the occurrence of power loss by the received signal, despite the short
distances of measurement, and these losses were more important in the NLOS con-
ditions compared to those observed in the LOS conditions. In [20], the authors in-
vestigated the accuracy of LoRa positioning using RSSI measured at the gateways,
and, in realistic conditions where the power attenuation caused by the radio link
was not known, the work reported ac-curacy errors of up to 588 m. Another recent
work [21], proposed an extensive position estimation algorithm to minimize the pos-
teriori RSSI error for multi-anchor cooperative estimation scenarios, and the results
showed that the location can be estimated with an accuracy less than 7 m; however,
this system was only tested in an outdoor scenario, which tends to be less chal-
lenging than indoor scenarios. Indeed, localization via the RSSI method is based
on signal power, which makes it very sensitive to multipath, and cannot provide
high accuracy information, as the correlation between the received power and the
distance is significantly influenced by the environment, which means that it cannot
reliably infer the target’s location information [22, 23].
In this chapter, the low-accuracy localization problem and the need for multiple
gateways or extensive algorithms to improve accuracy are overcome by integrating
UWB technology into the LoRa sensor; furthermore, real-time localization is enabled
thanks to the UWB’s high data rate and ability to detect tags whenever necessary, as
judged by the gateways. In this work, we propose a LoRa-UWB transceiver that
can be used as both a sensor-tag and an intermediate reader between the sensor-
tag and gateway, and this study mainly focuses on the localization rather than the
sensing function of the LoRa sensor-tag. The advantage of the proposed solution is
that the localization of the mobile sensor-tag is performed before sending anything
to the gateways (the distance between the sensor-tag and the UWB-LoRa reader is
computed at the reader node). This allows the comparison of the received location
information to the previous information, and we only send the data to the gateway
if they are different. Location via LoRa would first require directly sending the data
packet containing the time of the flight, then calculating the distance. In addition,
the possible inconvenience of the proposed solution mainly consists of deter-mining
how flexible the application is toward the ability to add the reader node between the
sensor-tag and the gateway.

6.3 UWB-LoRa localization method

This section describes, first, the transmission chain of the UWB-LoRa localization
scheme and then, the structure of the UWB-LoRa transceiver (reader and sensor-tag)
and its contents from modules to the required integrated UWB and LoRa antennas.

6.3.1 LoRa-UWB sensing and ranging approach

Figure 6.1 shows the working transmission chain related to the context in which the
LoRa-UWB transceiver is set to operate. As an example, the chain consists of several
6.3. UWB-LoRa localization method 105

LoRa sensors deployed in different locations that communicate with a LoRaWAN

gateway, as conventionally performed. In addition to these tags, a UWB- equipped
LoRa sensor-tag is placed and can directly communicate with either the gateway to
transmit sensing data, or an UWB-LoRa reader device to perform two-way ranging,
obtain the Time-of-Flight (ToF) at the reader node and compute the distance between
the two measures. The reader then sends the ranging information to the gateway
through LoRa signals. Since the gateway is knowledgeable a priori of the reader’s
location, it can determine the location of the sensor-tag and mon-itor its movement
in the environment.

F IGURE 6.1: Communication chain structure from the sensor tags to

the network and gateways.

Figure 6.2 presents the working organigram of the system and the roles of each
node in more detail. It consists of the LoRa gateway, a sensor-tag and a UWB-LoRa
reader that acts as an intermediate node between the two features.

F IGURE 6.2: Organigram of the working localization system high-

lighting the role of each node.
106Chapter 6. Multi-Standard UWB-LoRa High-Accuracy Long-Range Localization

In this setting, the gateway is required to determine the location of the sensor-tag,
especially if this location is moving in its environment. Following the ranging prin-
ciple using UWB technology, the reader sends an interrogation signal to the sensor-
tag, to which the tag replies by an acknowledgment signal. Through the measure of
ToF of these signals, the reader obtains a ranging estimate dn , which is the distance
between the sensor-tag and itself. While all these steps have been realized through
UWB protocol, the reader, then enters a decision phase: whether to send the obtained
location estimate to the LoRa gateway or not. The decision is taken after comparing
the actual distance value with the previous one a priori stocked. The reader will only
send the location of the sensor to the gateway if this sensor has moved and changed
location since the last measure. A microcontroller unit (MCU) assures the commu-
nication between the reader’s UWB and LoRa modules. If the new distance estimate
dn is not different from d(n−1) , then it is ignored, and the interrogation is sent again
to the sensor-tag. If it is different, then it means the tag has moved and its location
has changed. In this case, the estimate is sent to the LoRa module and automatically
from this module to the LoRa gateway through LoRa antennas, and dn is stocked by
replacing d(n−1) for the next loop iterations.
The main advantages of the presented localization technique are:
• The accuracy of the localization is of UWB-level accuracy which is approxi-
mately 15 to 20 cm and some cases even less. This accuracy cannot be achieved
with standard LoRa localization schemes, as LoRa is a narrow-band communi-
cation technology. At the receiving stage of the broadband time domain signal,
it is difficult to determine the arrival time precisely compared to UWB narrow
time pulses.
• The ability to achieve the real-time property of the localization, thanks to UWB
while still consuming as little power as possible. Real-time localiza-tion with
LoRa-only systems is not possible, as in this case, the LoRa module needs to
send the received packet from the tag to the gateway. This packet contains the
time of flight, which makes it necessary to estimate the location at the gateway
side each time, without knowing if the tag has moved or not, this continuous
packet transmission would consume more power which is against the LoRa
principle.
• UWB and LoRa based systems are mostly active (battery powered), in the pro-
posed system, the UWB battery powers both the UWB and LoRa modules.
Because LoRa’s power consumption is very low, another power unit is not
necessary.

6.3.2 Design of the LoRa-UWB transceiver

The designed UWB-LoRa transceiver is a multi-standard device capable of achiev-
ing the high-accuracy and long-range UWB localization of objects and sending the
lo-cation information to LoRa gateways. Its main purpose is to act either as a sensor-
tag or a reader intermediate node between the sensor-tag and the gateway. It per-
forms the localization of the tag and informs the gateway in case this latter has
moved in space.
The reader is composed of an UWB module (from STMicroelectronics [24]) and a
LoRa module, which communicate through an MCU unit. Hence, the reader has an
UWB antenna for RF communication with any UWB module and a LoRa antenna for
the RF communication with the LoRa gateways. The UWB module and the anten-
nas of the reader and tag’s working frequency channel is the UWB channel 2, which
6.3. UWB-LoRa localization method 107

starts at 3.75 GHz and ends at 4.25 GHz. The LoRa module and antennas work in
the 863 to 868 MHz band, which represents the Europe frequency channel for LoRa
communications, and it is part of the unlicensed Industrial, Scientific and Medical
(ISM) bands.
As illustrated in Figure 6.3, the reader board consists of a upper layer and a bottom
layer separated by an air gap. The upper layer consists of the radiating elements
of the UWB and LoRa antennas placed on the same FR4 substrate. In contrast, the
bottom layer consists of the ground plane and the LoRa module on one side, while
on the other side, an FR4 substrate with a branch-line coupler enabling the UWB cir-
cular polarization of the UWB antenna is placed, and at a distance from it, the MCU
and UWB module are integrated.

F IGURE 6.3: Structure of the UWB-LoRa transceiver, composed of

two layers, an electronics bottom layer and an antenna upper layer,
connected through signal vias.

Circular polarization was privileged for the UWB antenna, as it is more advanta-
geous due to it yielding better localization information and ranges compared to lin-
ear polarization; its advantages are described in detail in our previous works [14,25].
The common air gap between the two antennas serves as the air gap for standard
PIFA (Planar Inverted F-Antenna) design and simultaneously represents the neces-
sary gap between the substrate and ground, which guarantees a 500 MHz band-
width for the UWB probe-fed patch antenna.
Concerning the electronic part of the transceiver, the system complexity has been
kept at a low level during the design process. The main components are the power
supply unit, the MCU, the LoRa module and the UWB module. The first three are
mandatory and equivalent to any LoRa/LoRaWAN end-devices available on the
market, and the circuit connections between them are accordingly made to the con-
ventional systems. In addition, the required connections for the UWB module were
realized following the same model and firmware derived from the STMicroelectron-
ics UWB MEK1 localization board [20]. Finally, the UWB module is connected to the
MCU, which is also connected to the LoRa module.
Furthermore, the commercial cost of the LoRa module and MCU prevalent is low
(within 10 USD), and with the recent announcement of the STM32WL microcon-
troller from STMicroelectronics, the foreseeable cost will be even lower thanks to the
108Chapter 6. Multi-Standard UWB-LoRa High-Accuracy Long-Range Localization

integra-tion of the LoRa Module into the same die of the microcontroller. In contrast,
the costs of the UWB modules currently vary depending on their integration boards.
In this work, the module was provided as samples and are only commercialized
with the MEK1 board; however, other UWB modules are available at a relatively low
price (within 25 USD), such as modules from Decawave. To conclude, the proposed
concept of UWB-LoRa lo-calization can be applied and remains feasible, regardless
of the types of the LoRa or UWB modules chosen.

6.3.3 Antenna structure design

The structure is composed of two antennas (figures 6.4 and 6.5): a wideband patch
antenna operating at the UWB channel from 3.75 to 4.25 GHz and a PIFA antenna
operating at the LoRa channel from 863 to 870 MHz. The design and simulation were
realized using CST Microwave Studio software, where only the antenna parts of the
board were simulated, and the electronics modules positions were considered by
leaving the necessary space for their integration, which occurs later in the fabrication
process.
As the transceiver has two substrates, the radiating elements of both the UWB and
LoRa antennas are placed on the upper layer substrate. The bottom layer contains
other antenna elements: the ground plane shared by both antennas; the shorting pin
of the LoRa antenna, which extends until the ground plane; the feed vias; and the
branch-line coupler of the UWB antenna placed on the back surface of the bottom
substrate. Both substrates used are of the FR4 type and have a 0.8 mm thickness.

F IGURE 6.4: Antenna structure design consisting of a UWB patch an-

tenna and a LoRa PIFA antenna: (a1) top of the upper layer of the
board; (b1) bottom of the upper layer of the board; (a2) top of the bot-
tom layer of the board; (b2) bottom of the bottom layer of the board.

The UWB antenna design and optimization procedure is detailed in [26], it is a

probe-fed wideband patch using capacitive feeding from a small conductor printed
next to the radiating element. Since it is also designed to have circular polarization,
two orthogonal capacitive elements are designed. These two small rectangular ele-
ments are fed with vias originating from the coupler outputs placed at the bottom
layer. The dimensions of the UWB antenna (figure 6.4) are as follows: Substrate
6.3. UWB-LoRa localization method 109

F IGURE 6.5: Antenna structure design, consisting of a UWB patch

antenna and a LoRa PIFA antenna.

length and width are respectively Ls = 69 mm and Ws = 43 mm. The patch dimen-
sions are L p = Wp = 23 mm, designed with standard method to work at the center
frequency of 4 GHz. Each capacitive element’s length and width are lc = 10 mm,
wc = 1.5 mm, and the distance separating it from the radiating element is d=2 mm.
Finally, the classical branchline coupler is designed to operate at the same center fre-
quency of 4 GHz, from which its dimensions are derived.
The LoRa antenna is a slotted PIFA antenna. The slot starts from the feed point and
extends to help decrease the resonating frequency to the desired LoRa channel fre-
quencies. The external dimensions of the antenna (figure 6.4) are L = 14.5mm, and
its width is the same as the substrate width (Ws = 43mm). The slot width is 2.5
mm and its length dimensions are: L1 = 10 mm, L2 = 24.5 mm, L3 = 8 mm, and
L4 = 15 mm. The feed and short points are separated by a distance of 2.5 mm.
All antenna dimensions are summarized in Table 6.1.

UWB antenna Dimensions LoRa antenna Dimensions

Substrate length Ls 69 mm Conductor length L 14.5 mm
Substrate width Ws 43 mm Conductor width Ws 43 mm
Patch length and 23 mm Slot length l1 10 mm
width L p ,Wp
Capacitive element 10 mm Slot length l2 24.5 mm
length lc
Capacitive element 1.5 mm Slot length l3 8 mm
width wc
Capacitive element 2 mm Slot length l4 15 mm
separation from the
patch d
Coupler dimension 10 mm Slot width ws 2.5 mm
a
Coupler dimension 5.7 mm Separation between 2.4 mm
b feed and shorting
pin
Coupler dimension 2.87 mm Air gap (distance 7 mm
c from ground plane)

TABLE 6.1: Summary of the physical dimensions of the antennas.

As both antennas share the same ground plane, the air gap between this factor
110Chapter 6. Multi-Standard UWB-LoRa High-Accuracy Long-Range Localization

and the upper layer substrate is an optimized distance of 7 mm. The optimization
of this distance to accommodate both antennas is highly important as it significantly
affects the patch bandwidth and the PIFA antenna’s resonance performance.

6.4 Results and characterization

A prototype of the UWB-LoRa transceiver was realized and is presented in figure
6.6. The upper and bottom layers were realized separately and then were assem-
bled by soldering the feed connections from the coupler outputs to the capacitive
small patches next to the radiating element for the UWB antenna, and the feed and
shorting connec-tions from the ground plane to the upper layer PIFA element. The
transceiver board is powered by a small external battery and both antennas’ inputs
(LoRa and UWB) are powered by a signal via coming from their respective modules
(as shown in figure 6.3).

F IGURE 6.6: Prototype of the UWB-LoRa transceiver board: (a) the

upper layer of the board containing the ra-diating elements; (b) bot-
tom layer of the board containing the MCU, the modules and the
branch-line coupler.

6.4.1 Antenna measurements

Impedance matching
To measure the impedance matching of the antennas before testing the board, UFL-
type connections were designed a priori at both inputs of the branch-line coupler
(which represent the two input ports of the UWB antenna), and UFL-to-SMA cables
were used for measurements using a VNA (Vector Network Analyzer), as illustrated
in figure 6.7.
The SMA side of the cable connected to one of the coupler’s inputs is then loaded
by a 50 Ω match. In the case of the LoRa antenna, it can be seen that its feed was
connected directly to an SMA connector for the measurements (figure 6.7). Foam
with the same dielectric constant as air was also used between the two layers of the
board to facilitate the assembly of the board.
Figure 6.8 presents the simulated and measured reflection coefficient of the PIFA
antenna. Our results show a shift in frequency between the simulation and mea-
surement. This shift is due to the slot of length l4 , which was removed by covering
it with copper tape in the fabricated prototype, as shown in figures 6.6.a and 6.7.b.
In fact, the simulated result with the slot l4 shows that the antenna resonated at the
6.4. Results and characterization 111

F IGURE 6.7: Board assembly and measurements: (a) prototyping us-

ing foam to separate the layers and UFL connectors at the coupler
inputs; (b) stand-alone antenna matching measurement set up, using
a SMA connector for the LoRa antenna and UFL-to-SMA cables from
the coupler inputs for UWB antenna characterization.

desired frequencies; however, after the prototyping, it was observed that l4 lowered
the resonance frequency below 850 MHz, which is why it was covered. A slight dif-
ference in bandwidth is also observed, which is due to the reflection coefficient being
higher after adjusting the frequency, which led to the dampening of the bandwidth.
Finally, the measured result shows that the antenna’s impedance is matched at - 10
dB from around 845 to 875 MHz, which includes the desired LoRa channel (863 to
868 MHz) for LoRaWAN communication with the sensors and gateways.

F IGURE 6.8: Reflection coefficient of the LoRa PIFA antenna.

The measured and measured reflection coefficient of the UWB antenna is illus-
trated in figure 6.9. This coefficient was measured separately, via two measurements,
for both ports corresponding to the two inputs of the branch-line coupler by alter-
nating the isolation load between the two inputs. In both cases, one of the ports
must be isolated with a 50 Ω load. The advantage of choosing which port is the
input to the antenna is that the sense of the circular polarization can be chosen, that
is, left or right circular polarization can be obtained. The ability to choose the po-
larization sense implies that the antenna can be integrated into different systems
regardless of polarization specifications, such as in cases where, for example, the
112Chapter 6. Multi-Standard UWB-LoRa High-Accuracy Long-Range Localization

reader and tag antennas are required to have a specific polarization sense or the po-
larization of a reader (or tag) antenna is required to match the polarization of a tag
(or reader) antenna. A comparison between the simulation and measurement shows

F IGURE 6.9: Reflection coefficient of the UWB patch antenna for both
ports of the coupler (S11 and S22).

that the impedance matching of the antenna via simulation presents better results,
as it ranges between - 20 and - 28 dB, compared to - 12 and - 18 dB in the measure-
ment. This difference is due to factors such as the possible difference in air gap dur-
ing the assembly process, which can have a significant effect on the bandwidth and
impedance matching; another factor is the electronics modules and circuits, which
were not simulated with the antenna design. Finally, regardless of the port chosen,
our results show that the antenna’s measured reflection coefficient is matched at -
10 dB (from 3.55 GHz to 4.74 GHz), which includes the desired UWB channel (3.75
GHz to 4.25 GHz) in both cases. A slight difference between the two curves is ob-
served, which is due to the presence of the PIFA antenna, which is not completely
symmetrical with respect to the antenna axis.

Efficiency and gain

Further antenna measurements were realized in anechoic chamber to assess the gain
and radiation efficiencies of the antennas. Figures 6.10 and 6.11 show the simulated
and measured characteristics, respectively, for the LoRa and UWB antennas.
6.4. Results and characterization 113

F IGURE 6.10: LoRa antenna’s gain and efficiency characteristics.

The result of the measurement of the LoRa antenna shows a maximum efficiency
and gain at the operation frequency channel (863–870 MHz). The gain ranges be-
tween 0.82 and 0.91 dBi, while the efficiency is between - 1.65 and - 1.69 dB (cor-
responding to 68 % of the radiation). A shift in frequency is observed between the
measured and simulated curves for both the efficiency and gain, which is due to the
simulated design including the slot l4 , in contrast to the prototype, in which it was
removed to adjust the resonant frequency.
Furthermore, the UWB antenna’s simulated and measured efficiency and gain
are illustrated in figure 6.11. The results show stable gain and efficiency along the
desired frequency channel (3.75–4.25 GHz) in terms of both simulation and measure-
ment. Furthermore, the measured gain ranges between 5.1 dBi and 6 dBi on these
frequencies, where this variation of less than 1 dB is small enough to ensure that the
antenna behavior is considered identical at all the frequencies of the channel. Simi-
larly, the radiation efficiency varies between - 1.26 and - 1.4 dB, which corresponds
to radiation of between 72 and 75 %.

F IGURE 6.11: UWB antenna’s gain and efficiency characteristics.

114Chapter 6. Multi-Standard UWB-LoRa High-Accuracy Long-Range Localization

Radiation pattern
Figure 6.12 illustrates the measured radiation pattern of the UWB antenna, through
the realized gain of the co-polarization and cross-polarizations (here, the measure-
ments correspond to one arbitrary chosen port). The pattern is measured for the
UWB channel’s center and edge frequencies (3.75 GHz, 4 GHz, and 4.25 GHz) to
ensure the stability of its wideband operation.

F IGURE 6.12: Radiation pattern (azimuth) of the UWB antenna at the

channel’s center and edge frequencies, through polarization realized
gain.

As shown in the results, the antenna exhibits a 180° directional azimuth pattern
at all frequencies, and the circular polarization sense achieved is left hand (for the
measured port), with stable peak gains of 6 dBi.
Figures 6.13 and 6.15 illustrates the simulated polar and 3-dimensional radiation
patterns (realized gain) of the LoRa PIFA antenna around the transceiver board, at
a frequency of 868 MHz. As shown in these results, the antenna exhibits an omni-
directional radiation pattern and showcases two radiation dips along the PCB axis.
The antenna achieves measured peak gains between 0 and 1 dBi.
6.5. Discussion and experimental characterization of the board 115

F IGURE 6.13: Polar radiation pattern of the LoRa antenna at a fre-

quency of 868 MHz.

F IGURE 6.14: Three-Dimensional radiation pattern of the LoRa an-

tenna at 868 MHz frequency, with a view on both left and right sides
of the antenna structure.

6.5 Discussion and experimental characterization of the board

To assess the localization performance, two steps are investigated. The first one is
the UWB two-way ranging realized between the sensor-tag and the reader, and the
second one is the transmission of the location information to the LoRa gateway and
network. These steps are detailed in the following section.

6.5.1 From the Sensor-Tag to the Reader

Ranging measurements were performed using two prototypes of the UWB-LoRa
transceiver, with one used as a sensor-tag and another one used as a reader. It is
important to note that experimental tests with UWB signals are not prone to inter-
ference with other devices operating in the same environment [27, 28] because they
are spread over a wide band and transmitted at power spectral density levels close
116Chapter 6. Multi-Standard UWB-LoRa High-Accuracy Long-Range Localization

to the noise floor of conventional radio receivers, such as WiFi, Bluetooth, etc.
Figure 6.15 shows the set-up and environment of the measurements. An outdoor
scenario was privileged, as it enables more available space compared to buildings,
allowing us to measure the performance at LOS and analyze the maximum range of
detection that is achieved through UWB localization. This range does not de-scribe
the whole system’s UWB-LoRa range but only that of the UWB part of the system.
This maximum distance gives an idea of where the intermediate reader node needs
to be placed in such system, that is, how far from the sensor-tag it can be placed and
still be able to detect it.

F IGURE 6.15: Ranging with two LoRa-UWB transceivers, the first be-
ing a sensor-tag and the second being a reader.

Figure 6.16 illustrates the result of the ranging, through the Link Quality Indi-
cator (LQI) with respect to the distance measured via the reader (using its UWB
feature). The LQI is the ratio of the measured received signal strength to the already-
known saturation signal strength of the transceiver. It is a metric used by STMicro-
electronics for localization using their UWB module, which was integrated into the
designed transceiver.
6.5. Discussion and experimental characterization of the board 117

F IGURE 6.16: Ranging measurements of the UWB localization be-

tween the sensor-tag and the reader, in LOS and NLOS scenarios (out-
door).

Firstly, the results show that the LQI and, thus, the received power decreases
with respect to the distance, which is due to signal path loss. Indeed, having higher
levels of power is beneficial for obtaining longer ranges of detectability. For this rea-
son, solutions to overcome path loss exist in the literature and include the modeling
of the UWB propagation channel [29,30], the estimation of the channel path loss [31]
and the compensation for it at the signal processing level; however, the objective
in this work leans more toward demonstrating the feasibility of the UWB-LoRa lo-
calization concept by comparing it to conventional UWB and LoRa localization sys-
tems, the performances of which are known in terms typical range and accuracy.
Thus, the second observation from these results is that the reader can detect the mo-
bile sensor-tag up to distances of 120 m and 55 m in the LOS and NLOS scenarios.
The LQI naturally decreases with the increase in distance and presents small fluc-
tuations between close-distance measurements, as it depends on the environment
around the mobile tag at that time instant. The ranges achieved conform with typ-
ical maximum UWB ranges present in systems used in industry and the literature
(approximately 100–200 m for LOS).
It is necessary to mention that the LoRa feature of the sensor-tag in this work does
not have a role in the UWB localization process, which involves ToF measurements.
Indeed, its role is like that of any LoRa sensor, i.e., to allow the direct transmission
of sensing information to the gateway, which does not require sending it to the in-
termediate reader. Indeed, the intermediate reader and UWB communication part
of the sensor-tag only intervene if the sensor-tag has to be located via the gateway.

6.5.2 From the Reader to the Gateway and Network

After the ranging is performed, the reader node sends the location information of
the sensor-tag to the LoRa gateways, albeit only in case the latter has moved since
the last measurement. The reader transceiver compares the current obtained dis-
tance separating it from the sensor-tag and the previously computed distance. If
it is different, the reader sends the information through LoRa to the gateway. Fig-
ure 6.17 shows the information packet received by the network, and it displays the
118Chapter 6. Multi-Standard UWB-LoRa High-Accuracy Long-Range Localization

reader and the sensor-tag’s Identification Numbers (ID), the distance measured and
the LQI.

F IGURE 6.17: Ranging information of the sensor as displayed after its

reception via the LoRa network.

The results demonstrate that as long as the UWB-LoRa reader node is placed at
an appropriate distance from a LoRa-UWB (or UWB-only) sensor-tag, it is able to
localize through UWB high-accuracy ranging. The LoRa gateways placed at dis-
tances of kilometers from a sensor will be able to monitor this latter feature and stay
informed of its precise location whenever it changes in real time.

6.5.3 Localization accuracy

Localization using UWB is known to ensure the accuracy of the location information
that is of a sub-meter level. To observe this process, ranging measurements were re-
alized with the designed transceivers in an indoor scenario, where multipath effects
were often present, along a limited corridor of 35 m inside the LEAT laboratory. The
reader transceiver was placed at a fixed position, and the tag was moved, starting
from 10 to 35 m away from the reader. The measurements of the distance between
the reader and tag were performed for each step of 5 m. This operation was repeated
three times, and the measured distances were compared to the true distance between
the reader and tag. Figure 6.18 illustrates this comparison for all three measures.
The results show that for all three measures, the distance measured via the UWB
reader follows the reference curve (true distance) very closely. Measure 2 seems to
show the furthest measure from the reference curve. However, its accuracy remains
at the sub-meter level. Indeed, an example is shown at the 20 m distance, where
the UWB reader’s distance estimation is 19.67 m, which corresponds to an estimate
error of only 33 cm. All the other estimates (measure 1 and 3) achieve lower esti-
mate errors of between 10 and 30 cm. This result allows us to confirm the achieved
sub-meter accuracy due to UWB signals’ fine temporal characteristics (pulse width
of 2 ns). The achieved centimeter-level accuracy with UWB greatly benefits the
LoRa infrastructure; indeed, the conventional localization attempts using only the
LoRa technique to allow a gateway to locate a sensor, as shown in the literature, are
typically characterized by accuracy levels that exceed 10 m using Time-of-Flight
6.6. Conclusion 119

F IGURE 6.18: Comparison of UWB-measured distances with the true

distance between a reader and a tag in an indoor environment.

measurements, which may be low enough for outdoor approximate localization;

however, this accuracy level cannot accommodate indoor environments.

6.6 Conclusion
This chapter investigated the possibility of achieving high-accuracy, long-range lo-
calization through UWB and LoRa technologies. A multi-standard transceiver board
was proposed as a solution. It consists of a UWB module, a LoRa module and an
MCU, ensuring the communication between them. The transceiver can either be a
sensor-tag deployed in the LoRa infrastructure, a reader device, or both. It allows
the localization of a sensor-tag with UWB-level accuracy (a few centimeters) and the
transmission of the location information to LoRa gateways kilometers away from the
sensor-tag to the network. The UWB-LoRa transceiver was fabricated, and two pro-
totypes were used for ranging measurements in both outdoor and indoor scenarios.
The results demonstrated conformal UWB localization ranges and accuracy, as well
as their successful transmission to the LoRa gateway for the real-time monitoring of
the sensor, all while maintaining the low power requirements of the LoRa commu-
nications. In future work, the improvement in the proposed localization transceiver
could focus on the miniaturization aspect of the antennas, especially for the sensor-
tag, to help its integration into different objects that need to be tracked. The minia-
turization can be achieved by choosing a higher frequency channel and optimizing
the antennas accordingly. Another improvement aspect would be the integration of
direction detectability in addition to ranging, and this can be achieved via the de-
sign of an antenna array for the reader side, which can intercept the ranging signals
using the array elements and determine the angles of arrival related to the direction
component.
120 BIBLIOGRAPHY

Bibliography
[1] Khutsoane, O.; Isong, B.; Abu-Mahfouz, A.M. IoT devices and ap-
plications based on LoRa/LoRaWAN. In Proceedings of the IECON
2017—43rd Annual Conference of the IEEE Industrial Electronics So-
ciety, Beijing, China, 29 October–1 November 2017; pp. 6107–6112.
https://doi.org/10.1109/IECON.2017.8217061.

[2] Merhej, D.; Ahriz, I.; Garcia, S.; Terré, M. LoRa Based Indoor Local-
ization. In Proceedings of the 2022 IEEE 95th Vehicular Technology Con-
ference: (VTC2022-Spring), Helsinki, Finland, 19–22 June 2022; pp. 1–5.
https://doi.org/10.1109/VTC2022-Spring54318.2022.9860962.

[3] Marquez, L.E.; Calle, M. Understanding LoRa-Based Localization: Foun-

dations and Challenges. IEEE Internet Things J. 2023, 10, 11185–11198.
https://doi.org/10.1109/JIOT.2023.3248860.

[4] Mekki, K.; Bajic, E.; Chaxel, F.; Meyer, F. A comparative study of LP-
WAN technologies for large-scale IoT deployment. ICT Express 2019, 5, 1–7.
https://doi.org/10.1016/j.icte.2017.12.005.

[5] Augustin, A.; Yi, J.; Clausen, T.; Townsley, W.M. A Study of LoRa: Long Range
and Low Power Networks for the Internet of Things. Sensors 2016, 16, 1466.
https://doi.org/10.3390/s16091466.

[6] Devalal, S.; Karthikeyan, A. LoRa Technology—An Overview. In Proceedings

of the 2018 Second International Conference on Electronics, Communication
and Aerospace Technology (ICECA), Coimbatore, India, 29–31 March 2018; pp.
284–290. https://doi.org/10.1109/ICECA.2018.8474715.

[7] Muppala, R.; Navnit, A.; Devendra, D.; Matera, E.R.; Accettura, N.; Hus-
sain, A.M. Feasibility of Standalone TDoA-based Localization Using Lo-
RaWAN. In Proceedings of the 2021 International Conference on Localiza-
tion and GNSS (ICL-GNSS), Tampere, Finland, 1–3 June 2021 ; pp. 1–7.
https://doi.org/10.1109/ICL-GNSS51451.2021.9452311.

[8] Kombo, O.H.; Kumaran, S.; Bovim, A. Design and Application of a Low-
Cost, Low- Power, LoRa-GSM, IoT Enabled System for Monitoring of Ground-
water Resources With Energy Harvesting Integration. IEEE Access 2021, 9,
128417–128433. https://doi.org/10.1109/ACCESS.2021.3112519.

[9] Ferrero, F.; Truong, H.-N.-S.; Le-Quoc, H. Multi-harvesting solution for

autonomous sensing node based on LoRa technology. In Proceedings of
the 2017 International Conference on Advanced Technologies for Commu-
nications (ATC), Quy Nhon, Vi-etnam, 18–20 October 2017; pp. 250–253.
https://doi.org/10.1109/ATC.2017.8167627.

[10] Dardari, D.; Closas, P.; Djuric, P.M. Indoor Tracking: Theory, Meth-
ods, and Technologies. IEEE Trans. Veh. Technol. 2015, 64, 1263–1278.
https://doi.org/10.1109/TVT.2015.2403868.

[11] Dardari, D.; Conti, A.; Ferner, U.; Giorgetti, A.; Win, M.Z. Ranging with Ul-
trawide Bandwidth Signals in Multipath Envi-ronments. Proc. IEEE 2009, 97,
404–426. https://doi.org/10.1109/JPROC.2008.2008846.
BIBLIOGRAPHY 121

[12] IEEE Std 802.15.4z-2020; IEEE Standard for Low-Rate Wireless Net-
works–Amendment 1: En-hanced Ultra Wideband (UWB) Physical Layers
(PHYs) and Associated Ranging Techniques. IEEE: New York City, NY, USA,
2020. pp. 1–174.

[13] Coppens, D.; Shahid, A.; Lemey, S.; Marshall, B.V.H.C.; De Poorter,
E. An Overview of Ultra-WideBand (UWB) Standards and Organi-
zations (IEEE 802.15.4, FiRa, Apple): Interoperability Aspects and
Future Research Directions. IEEE Access 2022, 10, 70219–70241.
https://doi.org/10.1109/ACCESS.2022.3187410.

[14] 14. Benouakta, A.; Ferrero, F.; Lizzi, L.; Staraj, R. Antenna Characteristics Con-
tributions to the Improvement of UWB Real-Time Locating Systems’ Read-
ing Range and Multipath Mitigation. IEEE Access 2023, 11, 71449–71458.
https://doi.org/10.1109/ACCESS.2023.3294622.

[15] Pospisil, J.; Fujdiak, R.; Mikhaylov, K. Investigation of the Performance of

TDoA-Based Localization Over LoRaWAN in Theory and Practice. Sensors
2020, 20, 5464. https://doi.org/10.3390/s20195464.

[16] Daramouskas, I.; Mitroulias, D.; Perikos, I.; Paraskevas, D.; Kapoulas, V. Local-
ization in LoRa Networks Based on Time Difference of Arrival; Springer: Cham,
Switzerland, 2022; pp. 130–143. https://doi.org/10.1007/978-3-030-90528-6-12.

[17] Fargas, B.C.; Petersen, M.N. GPS-free geolocation using LoRa in

low-power WANs. In Proceedings of the 2017 Global Internet of
Things Summit (GIoTS), Geneva, Switzerland, 6–9 June 2017; pp. 1–6.
https://doi.org/10.1109/GIOTS.2017.8016251

[18] Ha, G.Y.; Seo, S.B.; Oh, H.S.; Jeon, W.S. LoRa ToA-Based Local-
ization Using Fingerprint Method. In Proceedings of the In-ternational
Conference on Information and Communication Technology Convergence
(ICTC), Jeju, Republic of Korea, 16–18 October 2019; pp. 349–353.
https://doi.org/10.1109/ICTC46691.2019.8939702.

[19] Aarif, L.; Tabaa, M.; Hachimi, H. Experimental test and performance of RSSI-
based indoor localization in LoRa Networks. Procedia Comput. Sci. 2022, 203,
420–425. https://doi.org/10.1016/j.procs.2022.07.055.

[20] Strzoda, A.; Marjasz, R.; Grochla, K. How Accurate is LoRa Positioning in Re-
alistic Conditions? In Proceedings of the Pro-ceedings of the 12th ACM In-
ternational Symposium on Design and Analysis of Intelligent Vehicular Net-
works and Applica-tions, Montreal, QC, Canada, 24–28 October 2022; As-
sociation for Computing Machinery: New York, NY, USA, 2022; pp. 31–35.
https://doi.org/10.1145/3551662.3561260.

[21] Li, B.; Xu, Y.; Liu, Y.; Shi, Z. LoRaWAPS: A Wide-Area Posi-
tioning System Based on LoRa Mesh. Appl. Sci. 2023, 13, 9501.
https://doi.org/10.3390/app13179501.

[22] Kwasme, H.; Ekin, S. RSSI-Based Localization Using Lo-

RaWAN Technology. IEEE Access 2019, 7, 99856–99866.
https://doi.org/10.1109/ACCESS.2019.2929212.
122 BIBLIOGRAPHY

[23] Lam, K.-H.; Cheung, C.-C.; Lee, W.-C. RSSI-Based LoRa Localization Systems
for Large-Scale Indoor and Outdoor Environ-ments. IEEE Trans. Veh. Technol.
2019, 68, 11778–11791. https://doi.org/10.1109/TVT.2019.2940272.

[24] B-UWB-MEK1—Evaluation kit for the B-UWB-MOD1 ultra-wideband mod-

ule—STMicroelectronics Available online: https://www.st.com/en/wireless-
connectivity/b-uwb-mek1.html (accessed on 3 February 2023).

[25] Benouakta, A.; Ferrero, F.; Lizzi, L.; Brochier, L.; Staraj, R. Measurements of
antenna polarization effects on Ultra-Wideband monitoring and localization.
In Proceedings of the 2021 IEEE Conference on Antenna Measurements and
Applications (CAMA); Antibes Juan-les-Pins, France, 15–17 November 2021;
pp. 589–590. https://doi.org/10.1109/CAMA49227.2021.9703597.

[26] Benouakta, A.; Ferrero, F.; Lizzi, L.; Staraj, R. Frequency Reconfigurable
and Circularly Polarized Patch Antenna Over Dual Ultra-wideband Chan-
nels. In Proceedings of the 2022 16th European Conference on Antennas
and Propagation (EuCAP); Madrid, Spain, 27 March–1 April 2022; pp. 1–5.
https://doi.org/10.23919/EuCAP53622.2022.9769006.

[27] Ghavami, M.; Michael, L.B.; Kohno, R. Ultra Wideband Signals and Systems
in Communication Engineering; 1st ed.; Wiley: Ho-boken, NJ, USA, 2007.
https://doi.org/10.1002/9780470060490.

[28] Duran, M.A.C.; D’Amico, A.A.; Dardari, D.; Rydström, M.; Sottile, F.; Ström,
E.G.; Taponecco, L. Chapter 3—Terrestrial Network-Based Positioning and
Navigation. In Satellite and Terrestrial Radio Positioning Techniques; Dardari,
D., Falletti, E., Luise, M., Eds.; Academic Press: Oxford, UK, 2012; pp. 75–153.
https://doi.org/10.1016/B978-0-12-382084-6.00003-9.

[29] Ghassemzadeh, S.; Greenstein, L.; Kavcic, A.; Sveinsson, T.; Tarokh, V. An Em-
pirical Indoor Path Loss Model for Ul-tra-Wideband Channels. J. Commun.
Netw.—JCN 2003, 5, 303–308. https://doi.org/10.1109/JCN.2003.6596612.

[30] Frattasi, S.; Rosa, F.D.; Dardari, D. Ultra-wideband Positioning and Tracking. In
Mobile Positioning and Tracking; Frattasi, S., Rosa, F.D. Eds.; Wiley: Hoboken,
NJ, USA, 2017. https://doi.org/10.1002/9781119068846.ch9.

[31] Rubio, L.; Reig, J.; Fernández, H.; Rodrigo-Peñarrocha, V.M. Experimental
UWB Propagation Channel Path Loss and Time-Dispersion Characterization
in a Laboratory Environment. Int. J. Antennas Propag. 2013, 2013, e350167.
https://doi.org/10.1155/2013/350167.
 123

General conclusion
The work realized in this thesis consisted in bringing enhancements to time-based
indoor localization, which is performed by real-time locating systems (RTLSs), which
employ Ultra-Wide Bandwidth (UWB) technology. The localization results obtained
with these systems call for improvements because they are mostly employed in com-
plex environments involving many mobile objects which makes precise localization
challenging. The principal enhancements and contributions realized in this work
focused on the conception of UWB antennas that are more optimized to the needs
of such RTLSs compared to the conventional antennas already integrated in most
commercial solutions.
The first chapter, presented an introduction to indoor localization, its techniques,
the error sources that can be encountered in different scenarios (line-of-sight and
non-line-of-sight). The technologies on which indoor localization is based were ex-
plained, with a focus on UWB technology by listing its advantages, applications,
regulations, standards, and consortiums. Then, a state of the art of the conventional
broadband and UWB antennas was realized. Finally, the old and recent require-
ments of UWB antennas were summarized and the recent expectations from UWB
antennas were highlighted.
In the second chapter, conforming to the recent IEEE standards specifying the emit-
ted UWB signal and its characteristics such as pulse width and pulse shape, a time-
domain study was realized through simulations by applying this signal to an UWB
transmitting antenna. The signal was generated by frequency upconversion to the
center frequency of the worldwide UWB channel centered at 7.9 GHz. The far-
electric-field received by field probes was analyzed in terms of pulse distortion, mag-
nitude, and time delay (or arrival). By extracting the time of arrival and comparing
it for different azimuth and elevation angles, it was confirmed that this time corre-
sponds in fact to the radial distance by considering the transmitting antenna as a
point source far from the receiver. It was also found that, while this time varies with
the distance, it does not vary with azimuth and elevation angles. This analysis helps
to conclude that the radiation pattern of an antenna does not affect the localization
results in time-based techniques such as UWB.
In the third chapter, a technical review of the UWB radio chips and evaluation boards
available in the industry was realized. The main UWB radio chips are from the man-
ufacturers : Qorvo, NXP, Apple and STMicroelectronics. Their characteristics and
the localization techniques they allow were summarized, some of these radio chips
allow not only ranging but also direction finding through the angle-of-arrival (AoA)
technique. Then, a focus on the antennas employed in these commercial systems
highlighted that most of the antennas are conventional UWB antennas, such as: lin-
early polarized monopole antenna, linearly polarized patch antenna or patch anten-
nas circularly polarized, however only at the center frequency of the UWB channel.
The chapter highlighted the influence of the chosen antennas on the localization re-
sults and that manufacturers always present the range and accuracy allowed by their
radio chips with conditions on the orientations of the reader and tag antennas. In
fact, the best localization result is always only achieved in a certain orientation of
the antennas. This discussion helps to conclude that by making the antennas more
efficient with higher gain and circular polarization, the localization will be improved
and most importantly will be robust to antenna orientations ensuring constant con-
fidence in the localization quality.
124 BIBLIOGRAPHY

In the fourth chapter, the conception of a frequency reconfigurable and circularly

polarized UWB antenna was realized. The antenna type was chosen to be patch for
higher gain compared to standard monopole UWB antenna. The bandwidth of the
patch was enhanced up to 500 MHz by suspending the substrate on air, and a capac-
itive feed mechanism was employed to compensate for the inductive effect brought
by the feeding probe. The frequency reconfigurability on two UWB channels (Chan-
nels centered at 4 GHz and 6.5 GHz) was achieved by modifiying the dimensions
of the radiating element by the means of copper connections which can eventually
be replaced by p-i-n diodes. The circular polarization was realized by the means
of two branchline couplers on a bottom substrate, each working at the center fre-
quency of one channel. The antenna was experimentally characterized and showed
good bandwidth and axial ratio performance.
In the fifth chapter, the previously designed antenna, working on the UWB channel
centered at 4 GHz, was tested in a commercial RTLS evaluation board from STMicro-
electronics and localization results achieved were compared to the results achieved
when using the commercial UWB antenna from the same manufacturer. Experi-
mental results obtained by ranging with a reader and tag allowed to demonstrate
the advantages brought by the designed antenna to the localization performance,
in terms of the improvement of : the reading range, the received power remaining
stable no matter the relative orientation of the reader and tag antennas, and in terms
of attenuation of the multi-path reflections which influence the received UWB signal
and thus the localization quality.
Finally, in the sixth chapter, a new concept was introduced to achieve both high-
precision and long-range localization by combining the two technologies : UWB
technology and Long-Range (LoRa) technology. Therefore, a multi-standard RTLS
transceiver board was designed and fabricated. The transceiver is composed of a
UWB module, a LoRa module, a microcontroller and of an antenna structure com-
posed of the previously designed UWB patch antenna and a planar inverted-F an-
tenna (PIFA) for LoRa operation at the 863 - 870 MHz channel. This transceiver was
placed as an intermediate node between a LoRa gateway and as a sensor which is
needs to be located. Experimental ranging was performed to assess the communica-
tion between the sensor, the reader (both represented by the same transceiver) and
the location information was sent to and received at the LoRa gateway and network.
 125

Perspectives
Building upon the work conducted in this thesis and its continuation, several
further developments and ways for improvement have been identified.

Concerning the time-domain analysis performed (chapter 2), while the time of
arrival was studied in function of the radiation pattern angles of the antenna, a
system-oriented study could be performed where the user could estimate the pa-
rameters of the localization system such as its reading range, or its accuracy in the
case of the presence of obstacles such as metal for example in the environment of the
antenna.

Concerning the frequency reconfigurable UWB antenna designed (chapter 4),

and which is meant for anchor systems, although its higher gain enables a longer
reading range, it was designed with the hypothesis that most anchor RTLSs are usu-
ally fixed on walls or ceilings, or placed on surfaces in indoor environments. Thus, if
the anchor were to be placed in a different location, it would be difficult to detect tags
if they are placed outside of its radiation pattern coverage. For this reason, comple-
mentary additions could be made by designing an anchor antenna with many facets
to allow for an omnidirectional radiation while still benefiting of the high gain of the
antenna.

Another possible development axis could be the study of direction-finding RTLSs

in addition to the range and localization RTLSs studied in this work. In fact, the
measurement of the direction of the tagged object is performed through the angle-
of-arrival (AoA) scheme, which requires the reader antenna to possess two or three
elements, that is, for this antenna to be an array antenna. In this case, the difference
between the times at which the signal from the tag arrives at these antenna elements
can be calculated and thus direction can be found.

Finally, to conclude, with the evolving potential of UWB technology and its pres-
ence being more and more pervasive in the most recent smartphones, cars and wear-
ables, it could be said that this technology has a promising future in scientific re-
search which could help its further development into a more standard communica-
tion scheme.
 127

List of Publications
• International journals :

1. A. Benouakta, F. Ferrero, L. Lizzi and R. Staraj, "Advancements in Industrial

RTLSs: A Technical Review of UWB Localization Devices Emphasizing Antennas for
Enhanced Accuracy and Range," MDPI Electronics 2024, 13, 751, doi: 10.3390/elec-
tronics13040751

2. A. Benouakta, T. M. Nguyen, F. Ferrero, L. Lizzi and R. Staraj, "Design of

a Multi-Standard UWB-LoRa Antenna Structure and Transceiver Board for High-
Accuracy and Long-Range Localization Applications," MDPI Electronics 2023, 12,
4487, doi: 10.3390/electronics12214487

3. A. Benouakta, F. Ferrero, L. Lizzi and R. Staraj, "Antenna Characteristics Con-

tributions to the Improvement of UWB Real-Time Locating Systems’ Reading Range
and Multipath Mitigation," in IEEE Access, vol. 11, pp. 71449-71458, 2023, doi:
10.1109/ACCESS.2023.3294622.

• International Conferences :

1. A. Benouakta, F. Ferrero, L. Lizzi and R. Staraj, "Frequency Reconfigurable

and Circularly Polarized Patch Antenna Over Dual Ultra-wideband Channels," 2022
16th European Conference on Antennas and Propagation (EuCAP), Madrid, Spain,
2022, pp. 1-5, doi: 10.23919/EuCAP53622.2022.9769006.

2. A. Benouakta, F. Ferrero, L. Lizzi, L. Brochier and R. Staraj, "Measurements of

antenna polarization effects on Ultra-Wideband monitoring and localization," 2021
IEEE Conference on Antenna Measurements Applications (CAMA), Antibes Juan-
les-Pins, France, 2021, pp. 589-590, doi: 10.1109/CAMA49227.2021.9703597.

• National Conferences :

1. A. Benouakta, F. Ferrero, L. Lizzi, R. Staraj, "Antenne Ancre Ultra-Large Bande

Reconfigurable en Fréquence pour applications de localisation et de positionnement
d’objets," Journées Nationale Microondes (JNM), Limoges, France, June 2022.

2. A. Benouakta, F. Ferrero, L. Lizzi, R. Staraj, "Antenne Ancre Ultra-Large Bande

Reconfigurable en Fréquence pour applications de localisation et de positionnement
d’objets," Conférence Plénière du GDR-Ondes, Lille, France, December 2021.

3. A. Benouakta, F. Ferrero, L. Lizzi, R. Staraj, "Conception d’une Structure An-

tennaire Multi-Standards ULB-LoRa Pour Applications de Localisation à Haute Pré-
cision et de Longue Portée," Journées Nationale Microondes (JNM), Antibes Juan-
Les-Pins, France, June 4 - 7th 2024.