Advances in Computer Science Research, volume 62

2nd Joint International Information Technology, Mechanical and Electronic Engineering Conference (JIMEC 2017)

A New Method To Measure The Noise Figure Of

Multi-Channel Radar Receiving System
Weimei Li, Bo Liu
Xi’an Institute of Space Radio Technology, CAST, Xi ’an 710100, Shaanxi, China

Keywords: noise figure, NF measurement, multi-channel A figure of merit and G/T for a multi-channel active array
radar. was derived by Lee[2] in terms of noise performance, and the
effects of the beam-forming network on the overall noise
Abstract figure of the system was also studied. A general formulation
for the computation of noise figure of active weighted array
Phased array radar or radar with array feed reflector antenna antennas is presented in [3]. An equivalent single-channel
is made up of multiple channels, and the radar detection range receiver model[4] was used to analyze the beam sensitivity of
is closely related to the noise figure (NF). Before the radar an antenna array receiving system, in which efficiencies and
being put into operation, it’s necessary to measure its NF to noise temperature are parameters of the equivalent single-port
evaluate the system performance. Here, an NF measurement antenna and equivalent amplifier. Two approaches[5,6] were
method of multi-channel radar receiving system is analyzed in recently introduced for defining the efficiencies and system
detail with the theory of equivalent conversion of a multi-port noise temperature of a receiving antenna array, one of which
network model into a two-port network model, combining is based on the isotropic noise response of the array and the
with channel characteristics of array antenna with various other on an equivalent system representation. However, none
amplitude weighting. To prove the correctness of the of the mentioned papers above involves the NF measurement
measurement method, a four-channel example is presented to method of a multi-channel radar receiving system. The
show that the equivalent conversion method is feasible, and it relationship between radar system noise figure and channel
also demonstrates that it is possible to precisely estimate NF noise figure is derived in [7], but the result is not applicable
of a multi-channel system through computation. for multi-channel radar receiving system with amplitude
weighting just like the radar with array feed reflector antenna.
1 Introduction This paper presents a new NF measurement method which
applies to every sort of the multi-channel radars. Here, as
Phased array radar or radar with array feed reflector antenna illustrated in Figure 1(a), a radar receiving system with array
occupied an increasingly important position in the modern feed reflector antenna is showed to represent the multi-
radar techniques. There usually exist complex feed network, channel receiving system, which consists of a number of
delay lines and transmit/receive (TR) module in the radar, and branches with different weighting amplitudes and phases, and
such radar receiving system consists of one or multi-stage the phased array radar receiving system can be viewed as
beam forming network together with the attenuator and the special case of Figure 1(a). To prove the correctness of the
phase shifter etc. Noise can enter the radar receiver together measurement method, the NF of the multi-channel radar
with the echo signal through antenna or be generated by the receiving system is derived using the basic formulas proposed
radar receiver itself. For a microwave radar receiver, the by [3].
external noise coming through the radar antenna is usually
small, so the sensitivity of the receiver is mainly determined 2 Noise characteristic analysis
by its inherent noise, which is also referred to as internal
noise. The noise level of radar receiver’s internal noise is The array feed elements of Figure 1(a) (such as horns) usually
characterized by NF, and the radar detection range is closely have the same gain, however the input powers and phases of
related to NF shown in radar equation[1]. the echo from the reflector of the antenna into array feed
In general, direct connection of an NF measuring instrument element are different, the phases of each branch after the
in the system under test can provide NF of a two-port receiver, action of the phaser reached the combiner are also different
but as for a radar receiving system with multiple channels, its unlike the phased array radar. All the phases of phased array
NF can hardly be measured directly due to the multiple input radar in the receiving channels are controlled in a phase plane
ports. Furthermore, the NF of multi-channel radar with array by the phase shifter, and the far field plane waves reached on
feed reflector antenna is rarely discussed in previous work. the radiating elements are also in the same amplitude. We can
For NF is the fundamental standard in designing a radar set the output powers of each array element in Figure 1(a) are
receiving system, it is necessary to compute the NF precisely. ( P1 , P2 Pn ), and not all of them are the same. For the radar
In the system trade-off study of a multi-channel radar, an
accurate measurement of the overall noise figure is also receiver excludes antenna in the common sense, the radar
necessary. receiving system in Figure 1(a) can be simplified into Figure

Copyright © 2017, the Authors. Published by Atlantis Press.

This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/). 252
 Advances in Computer Science Research, volume 62

1(b), each channel can be represented by three components, a figure of the single m th receiving branch before the
front-end attenuator (L1), an LNA, and a back-end attenuator, combiner, NFm can be written as ：
here, we divided back-end attenuator into two parts: the fixed
loss and the variable loss. The fixed loss includes ohmic loss NFA  1 L2 L3m  1
of the attenuator, the phase shifter, and cables etc. The
NFm  L1   (3)
1 L1 GA L1
variable loss is the differential weightings amplitudes denoted
by L3m. The LNA is characterized by a power gain GA and a The additive output noise power introduced by the combiner
noise figure NFA. The front-end loss includes the attenuation can be given as below:
K ( Ls  1)T0 B
in the circulator, limiter and cable losses etc. The ohmic loss
Ns  (4)
of the combiner is denoted by Ls. Note that all the loss factors Ls
are numerically greater than 1, and all the losses are truth
The total output noise power at the output of the combining
values, not in decibel in the following formula derivation.
networks is as below:
Based on the original definition of NF: n
N om
S N N S N
NF  in in  out  in  out 
1
(1) N ot    Ns (5)
Sout N out N in Sout KT0 B Gr m 1 nLs
The noise is incoherently summed(power sum), while the
where K is Boltzmann constant, T0 is the room temperature, signals in the power combiner are combined in phase, thus are
B is the noise bandwidth of the device, S in and S out are the summed coherently.
n
input and the output signal power of the device or a two-port The total input signal power S in in Figure 1(b) is Sin   Pm ,
m1
receiver, respectively. Nin  KT0 B is the input noise, and N out
and output signal power for m th receiving channel reaches
is the noise power at the output end of the receiver, Gr is the output of the combiner can be given as follows:
power gain of the two-port receiver, if the receiver has Som  PmGm / (nLs ) (6)
multiple-port receiving network, it represents equivalent Assuming that the final phase of each channel at the input end
receiving gain of the receiver.
of the combiner is  m after the phases shifting, the total
output signal power at the output of the combining networks
is given as follows:
P1 P2 Pn n 2

So   Som e jm
input loss

input loss

input loss

circulator (7)
L1 m 1
...
limiter

limiter

limiter

Thus we can get the equivalent receiving gain:

NFA,GA 2
 Pm  jm
NF
n


...
fixed loss
fixed loss

 e
fixed loss

Gr
LNA
L2
Pout So GA m 1  L3m 
Gr   (8)
attenuator

attenuator

attenuator

P
Sin nL1L2 Ls
variable loss

variable loss
variable loss

m
L3m
... m 1
phaser
So far, the receiving system in Figure 1(a) can be simplified
combiner LS combiner into Figure 1(c), with the equivalent gain Gr and noise figure
Pout Pout
(a) (b) (c)
of NF . Based on the definition of noise figure in Eq. (1), the
noise figure of the multi-channel radar receiving system is
Figure 1: Simplification of the receiving system with array expressed as follows:
feed reflector antenna 
n
1 
  KT B  NF
0 m  Gm 
nLs 
 Ns
The output noise power for each receiving channels before NF 
N ot 1
 
m 1  
1
the combiner can be given as follows: N in G r KT0 B Gr (9)
n

 nL L L  P
Nom  KT0 B  NFm  Gm
n m
1 L n 1
(2)   L1 NFA   1   1 2 s  m 1
2
 L
m 1 3m G L
A m 1 3m GA  n Pm  jθm
where Gm is the power gain of the m th receiving branch  
m 1 
e
L3m 
before the combiner, m  1, 2 n , and n is the number of
the branches for the multiple-port radar receiving system. So Note that, the power Pm in Eq. (9) does influence the NF, and
Gm  G A / ( L  L  L3m ) , from Figure 1 (b), NFm is the noise it represents the signal reflected from the antenna reflector.
1 2
The input signal of the receiving system is the far-field plane
wave from the scatterers. Because the reflector is part of the

253
 Advances in Computer Science Research, volume 62

receiving system, so the NF result is still not related to the n 2

input power. In the special case, this formula also applies to
the phased array radar receiving system. For a phased array
So '  
m 1
Som 'e jm '
(12)

radar, the input power P1  P2   Pn , and the phase at the where  m ' is the final modulated phase after phase shifting at
input end of the combiner is 1   2   n . the input end of combiner for each channel in Figure 2(a), and
the equivalent receiving gain can be got:
3 Noise figure measurement
2
n  Pm  jm '
 
m 1 
e
L3m 
Noise figure usually represents the performance of a two-port So ' GA
receiver, and its measurement is also based on a two-port Gr '   n
(13)
P
Sin ' nLg L1 L2 Ls
network, as for the radar receiving system as show in Figure m
1(a), the following will discuss its measurement method. m 1

Based on the analysis in section 2, a multi-port network Though the output signals after the power divider in Figure
shown in Figure 1(a) can be converted into a two-port 2(a) are in phase, and the phases at the input of each channel
network shown in Figure 1(c), an arbitrary power divider with in Figure 1(b) are different, nevertheless, the phases being
ohmic loss denoted as Lg can be added in the input end of the considered in the formulas are both the final results after
two-port network in Figure 1(c) , and the receiving system in phases shifting at each input end of combiner in the both
Figure 1(b) is turn into be what shown in Figure 2(a), all the figures, thus  m ' can be equal to  m .
component are the same as what in Figure 1(b) except for the
power divider.

L1 NFA,GA L2 L3m

input loss fixed loss variable loss

P1
combiner
divider

input loss fixed loss variable loss

P2
...

Pn input loss fixed loss variable loss

(a) Figure 3: Measurement of a four-branch equivalent
Lg NF,Gr experiment system
(b)

Figure 2: Equivalent measurement diagram of the receiving The total input noise power in Figure 2(a) is Nin '  KT0 B , and
system the noise power at the input end of channel marked in red
dashed line is N1  N 2   N m  KT0 B , whether the power
Based on the previous analysis, we can assume that the
system in Figure 2(a) with Lg=0 is just the system in Figure divider is arbitrary or equal, because the input noise power
1(b), that is NF of Figure 2(b) with Lg=0 is the same as that can be assumed as thermal noise, so they are same for all
of Figure 1(c) .To confirm this hypothesis, we compute the branches.
noise figure in Figure 2(a). The noise figure of the single m th receiving branch before
the combiner in Figure 2(a) can be written as ：
In Figure 2(a), set the input signal power is Sin ' , and the
L1  1 NFA  1 L2 L3m  1
signal power at the m th input port of each channel marked in NFm '  Lg    (14)
the dashed line is Pm , for the signals after the arbitrary power 1 / Lg 1 L1 Lg GA L1 Lg
n The output noise power for each receiving channels before
divider are different. The relationship of them is Sin '   Pm . the combiner can be given as follows:
m1
The power gain of the m th receiving branch, before the Nom '  KT0 B  NFm ' Gm ' (15)
combiner is: The additive output noise power introduced by the combiner
Gm '  GA / ( Lg  L1  L2  L3m ) (10) can be given as below:
K ( Ls  1)T0 B
Output signal power for m th receiving channels reaches the Ns  (16)
output of the combiner is as follows: Ls
Som '  PmGm '/ (nLs ) (11) The total output noise power at the output of the combining
networks in Figure 2(a) is as below:
The total output signal power at the output of the combining
networks is given as follows:

254
 Advances in Computer Science Research, volume 62

n
N om ' Acknowledgements
N ot '    Ns (17)
m 1 nLs This work was supported by Xi’an Institute of Space Radio
Substituting the Eq. (13) and Eq. (17) in to Eq. (1), we will Technology.
get the noise figure of the system as shown in Figure 2(a).
NF ' 
N ot ' 1
 
References
N in ' Gr '
[1] Skolnik, Merrill I., “Introduction to Radar Systems,” 3rd
n (18)
Lg   Pm Edition, McGraw Hill Inc., (2001).
 n
1 L n
1 nL1 L2 Ls  [2] Lee J J, “G/T and noise figure of active array antennas,”
 L1 NFA   1 L  
m 1

 m 1 L3m GA m 1 GA  n  Pm  jθm
2
IEEE Trans. Antennas Propagat., 41, pp. 241–244,

3m
 e (1993).
m 1  L3m 
[3] Gatti R V, Dionigi M, and Sorrentino R, “Computation
Comparing Eq. (9) with the Eq. (18), it is clearly that the NF of Gain, Noise Figure, and Third-Order Intercept of
results of measurement system in Figure 2(a) is Lg (if in Active Array Antennas,” IEEE Trans. Antennas
decibel) larger than the actual radar system for the Propagat., 52, pp. 3139–3142, (2004).
introduction of the power divider. To verify the correctness of [4] Ivashina M V, Maaskant R, and Woestenburg B,
our research result, a four-branch measurement model as “Equivalent system representation to model the beam
Figure 2(a) is built in Figure 3. The ohmic loss of the power sensitivity of receiving antenna arrays,” IEEE Antennas
divider is Lg =1dB, the S parameters of the divider Wireless Propag. Lett., 7, pp. 733–737, (2008).
are S 21 = S 41 =-6dB, S 31 =-4.5dB and S 51 =-8.5dB, L1=0.9dB, [5] Warnick K F and Jeffs B D, “Efficiencies and system
L2=8.3dB, Ls=0.4dB, GA=30dB, NFA=1.4dB, L3m=[31.5, 31.5, temperature for a beam forming array,” IEEE Antennas
10, 4.5]dB. In computation NF in Figure 3 by Eq. (18), the Wireless Propag. Lett., 7, pp. 565–568, (2008).
ohmic losses of attenuator and cables are included in L2, the [6] Warnick K F, Ivashina M V, Maaskant R, and
four weightings amplitudes from the attenuator are L3m. Woestenburg B, “Unified definitions of efficiencies and
system noise temperature for receiving antenna arrays,”
IEEE Trans. Antennas Propagat., 58, pp. 2121–2125,
NF ' Gr '
(2010).
Computation 7.792 5.676 [7] Lai H, Zhou J and Li B, “Noise figure measurement of
multichannel phased array receiver module,” 2011 IEEE
Measurement 7.67 5.71 4th International Symposium on Microwave, Antenna,
Propagation, and EMC Technologies for Wireless
Communications (MAPE), pp. 232-235, (2011).
Table 1: Computation and measurement result (dB).

It can be seen from Table 1 that computation and

measurement results from the equivalent measurement system
are basically in agreement, the little error is due to precision
of instrument and environmental influences of experiment.
The measurement data plus or minus the ohmic loss of the
power divider resulting the gain and NF of the multi-channel
radar receiving system based on previous analysis. So the NF
and gain of the four-branch radar receiving network as shown
in Figure 1(a) are to be 6.67dB and 6.71dB, respectively.

4 Conclusion
The theory of equivalent conversion of a multi-port
network model into a two-port network model, is used to
measure NF and gain of multi-channel radar receiving system,
and the NF and gain of the equivalent system are also
analyzed in detail to demonstrate the correctness of the
measurement method together with the measurement of a
four-way network. We can directly measure NF and gain of a
multi-port network with the help of a power divider, and the
measurement data plus or minus the ohmic loss of the power
divider resulting the gain and NF of the multi-channel radar
receiving system.

255