See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/317084775

Performance Analysis of LMS Adaptive Beamforming Algorithm

Article · July 2013

CITATION READS

1 215

4 authors, including:

Vivek Kumar
Università degli studi di Cagliari and Philips
25 PUBLICATIONS   88 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Implementation of Machine Learning and Deep Learning View project

Data Mining and Machine Learning for Knowledge Discovery and Predictive Analysis View project

All content following this page was uploaded by Vivek Kumar on 23 May 2017.

The user has requested enhancement of the downloaded file.

ISSN : 2230-7109 (Online) | ISSN : 2230-9543 (Print) IJECT Vol. 4, Issue Spl - 5, July - Sept 2013

Performance Analysis of LMS Adaptive

Beamforming Algorithm
1
Vivek Kumar, 2Deepak Rajouria, 3Manju Jain, 4Vikas Kumar
Dept. of ECE, LGOI, RGTU, Bhopal, India
1,3

2
SDO ( Transmission), BSNL, Bihar, Chhapra, India
4
Dept. of ECE, PIT, RGTU, Bhopal, India

Abstract The array factor for elements (Ne) equally spaced (d) linear array
The adoption of smart antenna techniques in future wireless is given by
systems is expected to have a significant impact on the efficient
use of the spectrum, the minimization of the cost of establishing
new wireless networks, the optimization of service quality, and AF (Φ) = (1)
realization of transparent operation across multi technology
wireless networks. where α is the inter element phase shift and is described as:
This paper presents brief account on Smart Antenna (SA) system
in context of adaptive beamforming particularly LMS. Further,
Implementation that resolves around the LMS adaptive algorithm, (2)
chosen for its computational simplicity and high stability into the and Φₒ is the desired direction of the beam.
MATLAB simulation of an adaptive array of a smart antenna base In reality, antennas are not smart; it is the digital signal processing,
station system, is to investigate its performance in the presence along with the antenna, which makes the system smart. When
of multipath components and multiple users. The capabilities of smart antenna is deployed in mobile communication using either
smart / adaptive antenna are easily employable to Cognitive Radio time division multiple access (TDMA) or code division multiple
and OFDMA system. access (CDMA) environment, exploiting time slot or assigning
different codes to different users respectively, it radiates beam
Keywords towards desired users only. Each beam becomes a channel, thus
Smart /Adaptive Antenna, Beam Forming, LMS, SDMA, avoiding interference in a cell. Because of these, each coded
OFDMA. channel reduces co-channel interference, due to the processing
gain of the system. The processing gain (PG) of the CDMA system
I. Introduction is described as:
Since Radio Frequency (RF) spectrum is limited and its efficient PG= 10log (B / Rb) (3)
use is only possible by employing smart/adaptive antenna array where B is the CDMA channel bandwidth and R b is the information
system to exploit mobile systems capabilities for data and voice rate in bits per second.
communication. The name smart refers to the signal processing If a single antenna is used for CDMA system, then this system
capability that forms vital part of the adaptive antenna system supports a maximum of 31 users. When an array of five elements
which controls the antenna pattern by updating a set of antenna is employed instead of single antenna, then capacity of CDMA
weights. Smart antenna, supported by signal processing capability, system can be increased more than four times. It can be further
directs narrow beam towards desired users but at the same time enhanced if array of more elements are used [2-7].
introduces null towards interferers, thus optimizing the service
quality and capacity. Consider a smart antenna system with (Ne) II. Beamforming
elements equally spaced (d) and user’s signal arrives from desired Beamforming is a general signal processing technique used to
angle 0 as shown in fig. 1 [1]. Adaptive beamforming scheme i.e. control the directionality of the reception or transmission of a
LMS is used to control weights adaptively to optimize Signal to signal on a transducer array. Beam forming creates the radiation
Noise Ratio (SNR) of the desired signal in look direction Φₒ. pattern of the antenna array by adding the phases of the signals in
the desired direction and by nulling the pattern in the unwanted
direction. The phases and amplitudes are adjusted to optimize the
received signal. A standard tool for analyzing the performance of
a beam-former is the response for a given N-by-1 weight vector
W (k) as function of, known as the beam response. This angular
response is computed for all possible angles.

III. Adaptive Beamforming

The adaptive algorithm used in the signal processing has a profound
effect on the performance of a Smart Antenna system. Although the
smart antenna system is sometimes called the ―Space Division
Multiple Access, it is not the antenna that is smart. The function
of an antenna is to convert electrical signals into electromagnetic
waves or vice versa but nothing else. The adaptive algorithm is the
one that gives a smart antenna system its intelligence. Without an
adaptive algorithm, the original signals can no longer be extracted.
Fig. 1: Smart Antenna Array System In the fixed weight beamforming approach the arrival angles does

w w w. i j e c t. o r g International Journal of Electronics & Communication Technology   47

IJECT Vol. 4, Issue Spl - 5, July - Sept 2013 ISSN : 2230-7109 (Online) | ISSN : 2230-9543 (Print)

not change with time, so the optimum weight would not need to A Uniform Linear Array (ULA) with N isotropic elements, which
be adjusted. However, if desired arrival angles change with time, forms the integral part of the adaptive beamforming system, is
it is necessary to devise an optimization scheme that operates shown [Fig-2]. The output of the antenna array is given by,
on-the-fly so as to keep recalculating the optimum array weight
that’s done by using adaptive beamforming algorithm. x(t)= s(t)a(θₒ) + + n(t)

s(t) denotes the desired signal arriving at angle θₒ and u(t) denotes
interfering signals arriving at angle of incidences θi respectively.
a(θₒ) and a(θi) represents the steering vectors for the desired signal
and interfering signals respectively. Therefore it is required to
construct the desired signal from the received signal amid the
interfering signal and additional noise n(t). The outputs of the
individual sensors are linearly combined after being scaled using
corresponding weights such that the antenna array pattern is
optimized to have maximum possible gain in the direction of the
desired signal and nulls in the direction of the interferers. The
weights here will be computed using LMS algorithm based on
Fig. 2: Block Diagram of Adaptive Beamforming Algorithm Minimum Squared Error (MSE) criterion. Therefore the spatial
filtering problem involves estimation of signal s(t) from the
The task of the algorithm in a Smart antenna system is to adjust received signal x(t)(i.e. the array output) by minimizing the error
the received signals so that the desired signals are extracted once between the reference signal d(t), which closely matches or has
the signals are combined. Various methods can be used in the some extent of correlation with the desired signal estimate and
implementation of an adaptive algorithm. the beamformer output y(t) (equal to wx(t)). This is a classical
In comparison, an human can even listen to sound which is weaker Weiner filtering problem for which the solution can be iteratively
than the interference. The adaptive algorithm in a smart antenna found using the LMS algorithm.
system serves a similar purpose as the brain in this analogy,
however it is less sophisticated. Our brain can perform the above V. LMS Algorithm Formulation
signal selection and suppression with only two ears, but multiple From the method of steepest descent, the weight vector equation
antennas are required for the adaptive algorithm so that enough is given by
information on the user signals can be acquired to perform the w(n+1)=w(n)+½µ[- (E{e2{(n)})]
task. In human beings, some people are more intelligent than Where μ is the step-size parameter and controls the convergence
others. In order for them to be more intelligent, they have to have characteristics of the LMS algorithm; e2(n) is the mean square
a more developed brain. Similarly, some algorithms are smarter error between the beamformer output y(n) and the reference signal
than other algorithms. A smart algorithm usually requires more which is given by,
resources than algorithms that are less intelligent. Unlike our e2 (n) = [d⃰ (n)–whx(n)]2
brain which is a free resource, more resources in the world of The gradient vector in the above weight update equation can be
technology always mean more expensive components and more computed as
complicated system.

IV. Least Mean Square Algorithm In the method of steepest descent the biggest problem is the
The Least Mean Square (LMS) algorithm, introduced by Widrow computation involved in finding the values r and R matrices in
and Hoff in 1959 [3, 9-10] is an adaptive algorithm, which uses real time. The LMS algorithm on the other hand simplifies this
a gradient-based method of steepest decent [8]. LMS algorithm by using the instantaneous values of covariance matrices r and R
uses the estimates of the gradient vector from the available data. instead of their actual values i.e.
LMS incorporates an iterative procedure that makes successive R(n) = x(n)xh(n)
corrections to the weight vector in the direction of the negative r(n) = d ⃰ (n)x(n)
of the gradient vector which eventually leads to the minimum Therefore the weight update can be given by the following
mean square error. equation,
w(n+1) = w(n) + μx(n)[d ⃰ (n) – xh(n)w(n) ]
= w(n) + μx(n)e ⃰ (n)
The LMS algorithm is initiated with an arbitrary value w (0)
for the weight vector at n=0. The successive corrections of the
weight vector eventually leads to the minimum value of the mean
squared error.Therefore the LMS algorithm can be summarized
in following equations; Output, y(n)= whx(n)
Error, e(n) = d ⃰ (n) – y(n)
Weight, w(n+1) = w(n) + μx(n)e ⃰ (n)
The LMS algorithm initiated with some arbitrary value for the
weight vector is seen to converge and stay stable for
0 < μ < 1/λmax
Where λmax is the largest eigen value of the correlation matrix
Fig. 2: LMS Adaptive Beamforming Network R. The convergence of the algorithm is inversely proportional

48 International Journal of Electronics & Communication Technology w w w. i j e c t. o r g

ISSN : 2230-7109 (Online) | ISSN : 2230-9543 (Print) IJECT Vol. 4, Issue Spl - 5, July - Sept 2013

to the eigenvalue spread of the correlation matrix R. When the

eigenvalues of R are widespread, convergence may be slow.
The eigenvalue spread of the correlation matrix is estimated
by computing the ratio of the largest eigenvalue to the smallest
eigenvalue of the matrix. If μ is chosen to be very small then the
algorithm converges very slowly. A large value of μ may lead to
a faster convergence but may be less stable around the minimum
value.

VI. Simulation Results for the LMS

Algorithm
Table 1: Parameters Used
Source Signal Angle 45º
Fig. 4: Polar and Rectangular Plot With Interference Signal at
Interference Signal Angle 60º 40º
Signal to Noise Ratio (SNR) 15dB
Signal to Interference Ratio (SIR) 3 dB

LMS Step-size Parameter 0.05

Optimized-LMS Step-size Parameter 0.05

RLS Forgetting Factor 0.999

A linear 8 element array is simulated in MATLAB to analyse the
results of LMS.
A simulation run of 100 iterations with parameters summarized in
TABLE 1 resulted in error plot as shown in fig. 3. LMS algorithm
exhibited a Brownian motion around minimum value. To analyse
the ability of algorithm to give maximum gain in the direction of Fig. 5: Polar and Rectangular Plot With Interference Signal at
source signal and placing the null in the direction of interference 60º
signal, simulations were performed using different interferer signal
directions for 500 iterations. Remaining all parameters was set
as mentioned earlier. Fig. 4, 5 and 6 shows polar and rectangular
plots of LMS algorithms for interference signal arrival angle of
40º, 60º and 90º respectively.
For comparing the dependency of algorithm on SNR and SIR, it
was compared at two values
• SNR=30dB, SIR=10dB
• SNR=30dB, SIR=10dB
Fig. 7 shows the behaviour of LMS algorithm respectively, for
mentioned SNR and SIR values.

Fig. 6: Polar and Rectangular Plot With Interference Signal at

90º

Fig. 3: Error Plot

Fig. 8: Polar and Rectangular of LMS Algorithm

w w w. i j e c t. o r g International Journal of Electronics & Communication Technology   49

IJECT Vol. 4, Issue Spl - 5, July - Sept 2013 ISSN : 2230-7109 (Online) | ISSN : 2230-9543 (Print)

VII. Conclusion [11] R. Martínez, A. Cacho, L. Haro, M. Calvo,“Comparative

The LMS algorithm is most commonly used adaptive algorithm study of LMS and RLS adaptive algorithms in the optimum
because of its simplicity and a reasonable performance. Since it combining of uplink W-CDMA”, IEEE, pp. 2258-2262,
is an iterative algorithm it can be used in a highly time-varying 2002.
signal environment. It has a stable and robust performance against [12] M. T. Islam, Z. A. Abdul Rashid, C. C. Ping,“Comparison
different signal conditions. However it may not have a really fast between non-blind and blind array algorithms for smart
convergence speed compared other complicated algorithms like antenna system”, CS-2006-1015, pp. 1-8, 2006.
the Recursive Least Square (RLS). It converges with slow speeds [13] R. S. Kawitkar, D. G. Wakde,“Smart antenna array analysis
when the environment yields a correlation matrix R possessing using LMS algorithm”, IEEE Int. Symposium on Microwave,
a large Eigen spread. Usually traffic conditions are not static, Antenna, Propagation and EMC Technologies for Wireless
the user and interferer locations and the signal environment are Communications, pp. 370-374, 2005.
varying with time, in which case the weights will not have enough [14] J. Razavilar, F. Rashid-Farrokhi, K. J. R. Liu,“Software radio
time to converge when adapted at an identical rate. That is, μ architecture with smart antennas: a tutorial on algorithms
the step-size needs to be varied in accordance with the varying and complexity”, IEEE Journal on Selected Areas in
traffic conditions. It was found that the performance of the LMS Communications, Vol. 17, No. 4, pp. 662-676, April 1999.
beamformer improves as more elements are used in the antenna
array. This improvement is seen in the form of sharper beams
directed towards the desired users. Using small or large spacing Vivek Kumar is currently a Lecturer
values could degrade the performance of the LMS beamformer. at Laxmipati Group of Institutions, in
It was also found that the performance improves as the angular Bhopal. Prior to this, he was working
separation between the incident signals increases. Moreover, it in power industry with KMA power
was noticed that increasing the number of elements of the antenna Solutions. He received B.E degree in
array ensures better performance towards signals with grazing Electronics and Comm. Engineering
incidence. from Rajeev Gandhi Technical
University in 2012. He published a
References number of papers in preferred Journals,
[1] LAL. C. GODARA,“Applications of Antenna Arrays to and worked on a range of projects of
Mobile Communications", Part I; Performance Improvement, Electronics and communication and
Feasibility, and System. closely related areas. His research activities are currently twofold:
[2] LAL. C. GODARA,“Applications of Antenna Arrays to while the first research activity is set to explore the wireless
Mobile Communications, Part II; Beam-Forming and communications, signal processing and adaptive antenna signal
Directional of Arrival Considerations”, Proceeding of the processing. The second research is focused on converging two
IEEE, Vol. 85, No. 8, pp. 1195-1245, August 1997. opposite paradigms-the Social (Humanities) and the Technological
[3] Simon Haykin, Adaptive Filter Theory, third edition (New (Engineering) on a common platform, enabling sustainable
Jersey, Prentice-Hall Inc., 1996). development.
[4] B. Widrow, S.D. Steams,"Adaptive Signal Processing ",
(New Jersey, Prentice-Hall Inc., 1985). Deepak Rajouria is presently S.D.E in
[5] Md. Bakhar, Dr. Vani R.M, Dr. P. V. Hunagund,“Eigen BSNL, India. He received B.E. degree
Structure Based Direction of Arrival Estimation Algorithms in Electronics and Comm. Engineering
for Smart Antenna Systems”, IJCSNS International Journal from DR. B. R. Ambedkar University,
of Computer Science and Network Security, Vol. 9, No. 11, Agra, Uttar Pradesh. He received
pp. 96-100, November 2009. his M-Tech. in Computer Science
[6] F. E. Fakoukakis, S. G. Diamantis, A. P. Orfanides, G. A. from IIT-Roorkee, Uttaranchal, in
kyriacou,“Development of an Adaptive and a Switched Beam year 2013. His research interests
Smart Antenna System for Wireless Communications”, include signal processing, Optical
Progress in Electromagnetics Research Symposium 2005, Fiber Communication, Mobile
Hangzhou, C hina, pp. 1-5, August 22-26, 2005. Communication, Broadband Technologies, Virtual Private
[7] Rameshwar Kawitkar,“Issues in Deploying Smart Antennas Network, Managed Leased Line Network and adaptive antenna
in Mobile Radio Networks”, Proceedings of World Academy signal processing.
of Science, Engineering and Technology Vol. 31, July 2008,
pp. 361-366, ISSN 1307-6884. Manju Jain is currently pursing
[8] Salvatore Bellofiore, Consfan fine A. Balanis, Jeffrey Foufz, masters in microwave engineering.
Andreas S. Spanias,"Smart-Antenna Systems for Mobile She received B.E. degree in Electronics
Communication Networks Part I: Overview and Antenna and Communication Engg. from JNCT,
Design", IEEE Antenna’s and Propagation Magazine, Vol. Rajeev Gandhi Technical University in
44, No. 3, June 2002. year 2009. She will receive her M-Tech.
[9] Bernard Widrow, Semuel D. Stearns,"Adaptive signal from R.K.D.F University, Bhopal in
processing", Pearson education Asia, Second Indian reprint, august, year 2013. Her research interests
2002. include digital communication, signal
[10] Frank Gross,"Smart Antenna for Wireless Communica-tion", processing, and adaptive antenna signal
Mcgraw-hill, September 14, 2005. processing.

50 International Journal of Electronics & Communication Technology w w w. i j e c t. o r g

ISSN : 2230-7109 (Online) | ISSN : 2230-9543 (Print) IJECT Vol. 4, Issue Spl - 5, July - Sept 2013

Vikas Kumar is a final year Engineering

student of Electronics and Comm.
Department, Patel Institute of Technology
at Rajeev Gandhi Technical University,
Bhopal. His Research interests include
signal processing, wireless and optical
fiber communication.

w w w. i j e c t. o r g International Journal of Electronics & Communication Technology   51

View publication stats