EITN90 Radar and Remote Sensing
Lecture 8: Radar antennas
Daniel Sjöberg
Department of Electrical and Information Technology
Spring 2020
Outline
1 Basic antenna concepts
2 Effect of the antenna on radar performance
3 Reflector antennas
4 Phased array antennas
5 Array architectures
6 Conclusions
2 / 56
Learning outcomes of this lecture
In this lecture we will
I Review basic antenna concepts, particularly gain
I How the antenna affects the radar application
I Study the two main high-gain antenna solutions:
I Reflector antennas
I Phased array antennas
Target
Transmit signal Antenna
Transmitter T/R
Receive
signal
Receiver
protector
switch Mixer/Preamplifier
Detection and
Signal
Detector A/D measurement
processor
results
Low noise Local
IF amplifier
amplifier oscillator
Receiver (Adapted from Fig. 1-1)
3 / 56
Outline
1 Basic antenna concepts
2 Effect of the antenna on radar performance
3 Reflector antennas
4 Phased array antennas
5 Array architectures
6 Conclusions
4 / 56
The isotropic antenna
Pt
Radiation intensity: I= [W/steradians]
4π
Pt
Power density: Qt = [W/m2 ]
4πR2
5 / 56
Angular selectivity of an array
The received complex amplitudes may sum up constructively
(normal incidence θ = 0, solid arrows), or destructively (oblique
incidence θ > 0, dashed arrows). The result is angular selectivity.
6 / 56
Radiation pattern
αλ
Beam width: θ3 = , α≈1
L
Note that many results in the book are valid for large antennas,
but not necessarily for small (where antenna dimension L λ). 7 / 56
Directivity, gain, and efficiency
The directivity and gain of an antenna are both measures of
radiation intensity I, but are differently normalized:
def I(θ, φ)
D(θ, φ) = , Prad = power radiated from antenna
Prad /(4π)
def I(θ, φ)
G(θ, φ) = , Pacc = power accepted by the antenna
Pacc /(4π)
The powers are related by Prad = Pacc − Ploss , where Ploss is the
power lost in the antenna, for instance resistive losses. Thus,
G ≤ D, and the antenna radiation efficiency is η = G/D.
ηa 4πA 4π(0.88)2 32 000
Dmax = 2
≈ =
λ θ3 φ 3 θ3 [degrees] φ3 [degrees]
The above applies to a uniformly illuminated rectangular antenna.
The aperture efficiency is ηa .
8 / 56
IEEE standard for definition of terms for antennas
9 / 56
Sidelobes
Radiation in sidelobes is usually undesired. It is characterized by
the maximum sidelobe level relative the main beam (dB) or
relative an isotropic antenna (dBi).
Dmax = 43 dBi, SLL = −36 dB, or SLL = 7 dBi.
10 / 56
Average sidelobe level
Another characterization of sidelobe radiation is the average ratio
of sidelobe power to that of an isotropic antenna with the same
input power:
PSL
ΩSL Pt − PMB 4π 1 − PPMB
t
PMB
SLLave = Pt
= = ΩMB
≈1−
4π
4π − ΩMB Pt 1 − 4π Pt
I Pt = total radiated power
I PMB = power radiated in main beam
I PSL = power radiated in sidelobes
I ΩMB = main beam solid angle (typically ΩMB 4π)
I ΩSL = sidelobe solid angle
The simplified version can be interpreted as power conservation.
Discussion
11 / 56
Aperture tapers
The sidelobe structure can be controlled by the spatial distribution
of the electric field across the aperture, either by shaping a
reflector or controlling the elements of an array.
12 / 56
Example: Taylor tapering
13 / 56
Example: Taylor tapering
For lower side lobe levels, the beamwidth is increased.
Very low SLL:s, below −40 dB, are very difficult to realize, due to
finite tolerances in antenna components, reflector shape, thermal
effects, antenna alignment etc.
14 / 56
Phase errors
De 2 δrms = 5◦ ⇔ 3605 1
= 72 fraction of λ, requiring
= exp(−δrms )
D0 control on scale 0.42 mm at 10 GHz.
15 / 56
Outline
1 Basic antenna concepts
2 Effect of the antenna on radar performance
3 Reflector antennas
4 Phased array antennas
5 Array architectures
6 Conclusions
16 / 56
Applications
The antenna has a very direct effect on the radar system.
I The radar can only see within the antenna’s field of view
(FOV).
I The maximum range is limited by the gain of the antenna:
1/4
Pt G2 λ2 σ
Rmax =
(4π)3 Ls Pmin
I The spatial resolution ∆R = c/(2B) depends on bandwidth
B, which is often restricted by the antenna function.
I Different combinations of average transmit power and
effective area may be important, Pave A2e and Pave Ae :
Pave A2e SNR · 4πkT0 F Ls R4 · PRF
Track =
λ2 σ
SNR · 4πkT0 F Ls R4 Ω
Search Pavg Ae =
σ Tfs
17 / 56
Constant track/search performance
Discussion
18 / 56
Monopulse
A target’s angular location can be accurately determined in one
pulse (no sweep) by using two closely spaced beams in the antenna.
|∆(θ)|
verror (θ) = cos β β = arg(∆/Σ)
|Σ(θ)|
19 / 56
Monopulse — increased angular sensitivity
s 2
θ km θ
∆θ = √ 3 1+ km = slope of curve
km 2 · SNR θ3
20 / 56
Monopulse
Transmit with Σ pattern, receive in both Σ and ∆. Can add the
same functionality in elevation at the cost of more antenna ports.
21 / 56
Outline
1 Basic antenna concepts
2 Effect of the antenna on radar performance
3 Reflector antennas
4 Phased array antennas
5 Array architectures
6 Conclusions
22 / 56
Reflector antennas
Often rotational symmetric. Note the subreflector (or antenna
feed) needs to be supported in the aperture.
23 / 56
Parabolic reflector
The reflector is often characterized by the focal length to diameter
ratio, f /D.
24 / 56
Feed
The higher f /D, the more directive feed horn is necessary to avoid
spillover (power that misses the reflector).
25 / 56
Subreflector (Cassegrain configuration)
The size requirements can be relaxed by using a subreflector.
26 / 56
Offset feed
27 / 56
Outline
1 Basic antenna concepts
2 Effect of the antenna on radar performance
3 Reflector antennas
4 Phased array antennas
5 Array architectures
6 Conclusions
28 / 56
Phased array
With a phased array, the antenna beam can be scanned
electronically. Much faster than mechanical steering, but more
complex electronic implementation.
29 / 56
Array factor
N
1 X 2π
AF(θ) = exp −j n∆x sin θ − φn
N λ
n=1
Discussion
30 / 56
2π
Uniform phase shift: φn = λ n∆x sin θs
N
1 X 2π
AF(θ) = exp −j n∆x(sin θ − sin θs )
N λ
n=1 31 / 56
Phase shifters
The phase shift in each antenna element can be controlled by a
phase shifter.
32 / 56
Phase shifters, quantization error
More bits give higher resolution and less gain loss, but also higher
insertion loss (there are losses in each stage of the phase shifter).
Phase shifters are typically narrow-band.
33 / 56
Grating lobes
From the array factor
N
1 X 2π
AF(θ) = exp −j n∆x(sin θ − sin θs )
N λ
n=1
it is seen that it is maximized when
∆x
(sin θ − sin θs ) = 0, ±1, ±2, . . .
λ
The zero is the intended main beam θ = θs , but with large enough
spacing ∆x other angles θ can correspond to the non-zero
integers. These are called grating lobes.
If scanning is restricted to the region θ ∈ [−θs , θs ], grating lobes
are absent if
λ
∆x ≤
1 + | sin θs |
For ∆x < λ/2, there are no grating lobes regardless of θs .
34 / 56
Grating lobes, example for ∆x = λ
35 / 56
Grating lobes
36 / 56
Gain loss
When scanning electrically, the physical aperture has a fixed
orientation. The width of the main beam (directivity) is
determined by the projected aperture. Typically scanning with a
planar array would be constrained to ±60◦ or less.
37 / 56
Array element pattern
N
Ee (θ) X 2π
E(θ) = Ee (θ) AF(θ) = exp −j n∆x(sin θ − sin θs )
N λ
n=1
38 / 56
Array element influence on beam steering
39 / 56
Typical effect when scanning
Scan region
40 / 56
Wideband phased arrays
The narrow bandwidth of phase shifters may cause the beam to
point in different angles depending on frequency. The tolerable
θ3
fractional instantaneous bandwidth is Bi = 2 sin θs . Problem for
wideband radars with narrow beamwidth.
41 / 56
Wideband phased arrays
Instead of phase shifting, the time delay unit (TDU) implements a
fixed time delay in each chain. However, they have long been
considered bulky and costly. Discussion 42 / 56
Wideband phased arrays: subarrays
43 / 56
Outline
1 Basic antenna concepts
2 Effect of the antenna on radar performance
3 Reflector antennas
4 Phased array antennas
5 Array architectures
6 Conclusions
44 / 56
Passive array architecture
High power handling in all array components.
45 / 56
Active array architecture
Individual T/R units behind each antenna element, phase and
amplitude can be controlled. Costly, but extremely flexible.
46 / 56
Trade-off power-aperture
Power requirements in each element can be traded with aperture
size: low power requires larger area and more elements, but each
element may be cheap and there is less need for cooling.
47 / 56
Subarray architecture
Breaking down the array into subarrays may simplify
manufacturing and maintainance. Digitizing the signal at subarray
level allows many simultaneous beams (achieved by processing).
48 / 56
Simultaneous beam operation
Since the centers of the subarrays are far apart, grating lobes
appear when scanning the subarray AF, but are suppressed by the
fixed subarray pattern.
49 / 56
Overlapped subarrays
The subarray distance can be reduced by overlapping subarrays, at
the price of a more complicated feeding network (since each
element will be connected to more than one subarray).
50 / 56
Outline
1 Basic antenna concepts
2 Effect of the antenna on radar performance
3 Reflector antennas
4 Phased array antennas
5 Array architectures
6 Conclusions
51 / 56
Conclusions
I Basic antenna concepts: directive radiation pattern, side
lobes, aperture tapering.
I Antenna effects on radar function: FOV, gain and range,
power-aperture tradeoff.
I Improved angular localization with sum and difference
patterns.
I Reflector antennas provide high gain in different
configurations, but little scan possibility.
I Phased arrays are extremely flexible, but also costly. Various
levels of analog/digital solutions.
52 / 56
Discussion
Consider an antenna where all sidelobes have similar levels around
SLL = −30 dB. Using the relation
PMB
SLLave ≈ 1 −
Pt
provide an estimate of how much of the total power Pt is
contained in the sidelobes.
Go back
53 / 56
Discussion
Consider an antenna where all sidelobes have similar levels around
SLL = −30 dB. Using the relation
PMB
SLLave ≈ 1 −
Pt
provide an estimate of how much of the total power Pt is
contained in the sidelobes.
PSL Pt −PMB PMB
Answer: Pt = Pt =1− Pt ≈ −30 dB = 10−3 .
Go back
53 / 56
Discussion
With a linear cost model (unrealistic!) of 1 000 SEK/m2 and
1 000 SEK/kW, what would you recommend for a search radar
application, large aperture area, high power, or something in
between?
Go back 54 / 56
Discussion
With a linear cost model (unrealistic!) of 1 000 SEK/m2 and
1 000 SEK/kW, what would you recommend for a search radar
application, large aperture area, high power, or something in
between?
P
Answer: Cost = ( mA2 + kW ) · 1 000 SEK. For fixed performance,
the cost is dominated by either aperture area or power when one
becomes large. The lowest cost is in between. Go back 54 / 56
Discussion
N
1 X 2π
AF(θ) = exp −j n∆x sin θ − φn
N λ
n=1
How should the phase in each element, φn , be chosen to maximize
radiation in direction θ = π/4?
Go back
55 / 56
Discussion
N
1 X 2π
AF(θ) = exp −j n∆x sin θ − φn
N λ
n=1
How should the phase in each element, φn , be chosen to maximize
radiation in direction θ = π/4?
Answer: The exponentials add in phase (argument is zero) at
θ = π/4 if φn = 2π
λ n∆x sin(π/4). For a more general result,
replace π/4 by an arbitrary scan angle θs .
Go back
55 / 56
Discussion
A possible implementation of time delay units is to change length
of a transmission line leading up to the element, for instance with
a mechanical trombone structure as below.
Why would this be considered bulky?
Go back
56 / 56
Discussion
A possible implementation of time delay units is to change length
of a transmission line leading up to the element, for instance with
a mechanical trombone structure as below.
Why would this be considered bulky?
Answer: The time delay is proportional to the length of
transmission line used, which has to be significantly meandered to
fit in a small space if a long delay is required.
Go back
56 / 56
