Research on Simulation of Radar Return

from Extended Target

Xiaohui Yu, Pengfei Luo
School of Electronic Science and Engineering
National University of Defense and Technology
Changsha, China
[email protected]

AbstractSimulation of radar return from targets is an

important part of radar system simulation and it is developing
fast these years. Although there are already many mature
theories on simulation of radar return from targets, these
theories mainly aim at point targets and often simply describe an
extended target as synthesis of many point targets. There is little
theory to describe detailed methods of how to judge the position
of extended targets compared to the beam of radar and how to
divide extended targets into point targets and synthesize radar
returns from them. This thesis found the methods, described
them in detail and proved them with simulation. Using these
methods, we can simulate radar return of extended targets more
veraciously and provide more reliable data for analysis,
designation and test of radar systems.
Keywords-system simulation ;radar return; extended target;
minimal resolution cell

I.

the Doppler frequency shift; and , the elevation and the

azimuth between the line from radar to target and the axis of
the beam.
The delay can be ignored and the video signal received
can be described as
1/2

INTRODUCTION

Modern radar systems are becoming more and more

complex and expensive as in [3]. To reduce cost in research,
design and test of radar systems, radar simulation is rapidly
gaining in popularity. With this technology, performance of
radar systems can be ensured with fewer expenses, which made
it one important method of analysis, designation, test,
evaluation, battle training and so on of radar systems. The
simulation of radar return from targets is an important part of
radar system simulation, which already has many mature
theories with the development of radar simulation as in [4]. In
these theories, targets are mainly be modeled as point targets.
However, this hypothesis is unsuitable under some conditions
such as seeker radar which is low to level and the projection of
its beam onto level is little and some targets such as large ships
cannot be modeled as point targets compared to it. Thus, these
targets should be modeled as extended targets such as line or
surface targets. Then the problem of how to divide extended
targets and synthesize the radar return is prompt. In this thesis,
we discussed the method of simulating radar return of extended
targets under the condition of seeker radar of anti-ship missile.
II.

1/2

2
(1)
SR (t) = ST[t ]ej2( fo + fd )(t ) 3 4 G(,)

(4
)
R
L

Where ST (t ) is the signal to be transmitted; ,the

round-trip delay of the signal; R , the distance from radar to
target; L , the loss factor when the signal transmitting in radar
and the air; G , the one-way antenna power gain in the
direction of target; , the wavelength of the signal; , the
complex reflection coefficient; f o , the carrier frequency; fd ,

RADAR RETURN FROM POINT TARGETS

As in [6], the model of radar return from point target is

described as

2
(2)
Sv (t) =
G u(t ) e j 2 fd t
3 4
(4 ) R L
From (2) we can see, the distance R , Doppler frequency
shift fd , antenna gain G , complex reflection coefficient

and complex envelop of the transmitted

u (t ) ( ST (t ) = u (t )e j 2 f t decide the radar return.

signal

III.

THE FLOW OF SIMULATION

The flow chart of simulating radar return of extended

targets is described in Fig. 1. Each step will be explained in
detail below.
A. Getting Parameters and Judging the Position of Extended
Targets
A reference frame should be built with taking the center
point of the area concerned as the origin and the sea level as the
x-y surface, axis-z pointing above, axis-x pointing to the east.
The position information of the radar and the targets such as
longitude, latitude and height should be transformed into the
reference frame. Lets suppose the head of the target to be at
( xheadn , yheadn , 0) , the end of the target at ( xendn , yendn , 0) , the seeker
radar at ( xRadar , y Radar , z Radar ) , the elevation and the azimuth to be

978-1-4244-4994-1/09/$25.00 2009 IEEE

find if the head or the end of the target is in the beam, then the
part of the target from this point (head or end) to the point of
intersection between the targets head and end is the range in
the beam.
Radar

R1
R
Range

Target

dR

Cell 1

Figure 2. Dividing into minimal resolution cells

B. Dividing into Minimal Resolution Cells

If targets are in the side lobe of the beam and there is power
of target echo in each return, the target echo is faintness. Only
when the target is in the main lobe of the beam, strong echo
Figure 1. Simulation flow
can be received as in [7]. So we should only take the return of
Antenna and Antenna and the width of the beam to be 2 3dB .
main lobe into account and neednt divide the target in
The center of the beam projection which is also an ellipse is
azimuth. According to the shape of ships, they can be modeled
targets. The distance
as line
x0 =xRadar +(tan(Antenna+3dB)+tan(Antenna3dB)zRadar cosAntenna/2
3

resolution is decided by the transmit signal. Suppose the

y0 =yRadar +(tan(Antenna+3dB)+tan(Antenna3dB)zRadarsinAntenna/2
resolution of the radar to be dR , and then the target can be
divided into N parts:
Take the center as origin and the major axis of the ellipse as
x-axis to build another reference frame. X-axis points to the
7
N = Range
dR
direction of the beam. Then the ellipse equation is
4x2
+
Look at Fig. 2. Line R is from radar to the end of the target
(tan(Antenna +3dB)tan(Antenna 3dB)2zR2adar
4 in the beam and
Range is the part in Line R whose projection
2y2
is
the
targets
projection
in the major axis of the ellipse.
=1
(tan(Antenna +3dB)tan(Antenna 3dB))zR2adar tan3dB
Shorten Line R at dR and get line R1.Find the point of
intersection of line R1 and the major axis of the ellipse, then
the part from this point to the end of the target in the beam is
Transform the head and end coordinates of the target into
the first resolution cell. Do like above and find the point of
this reference frame:
intersection of line R2 which is Line R shortened at 2dR and
xheadnto = xheadn x0
the major axis of the ellipse. The second resolution cell is
y
5
headnto = yheadn y0
divided out. Do so until the Nth resolution cell is divided out.

Take the ends or the midpoints of these resolution cells as point

xendnto = xendn x0
yendnto = yendn y0
targets.
C. Getting the Antenna Gain
And taking the target as a line, its equation is
The antenna gain is the measurement of the ability of the
antenna
to transmit and receive power in given direction. Its
x xheadnto
y yheadnto
6 often described as antenna pattern. The amplitude of the signal
=
xendnto xheadnto yendnto yheadnto
received by radar is modulated by the antenna gain. The
The ellipse equation and line equation can be used to judge
if the target is in the beam of radar and get the range of the
target in the beam as in [5]. Find if the ellipse and the line have
point of intersection (Equal to find the solution of (4) and (6)).
If there is no point of intersection or only one, the target must
not be in the beam. If there are two points of intersection, find
if the two are between the targets head and end. If they arent,
the target must not be in the beam. If both of them are, the part
between the two points of the target is the range in the beam.
Else, if only one of them is between the targets head and end,

antenna pattern function is the antenna gain decided by the

elevation and the azimuth between the line from radar to target
and the axis of the beam. In the simulation of radar system, the
antenna pattern function can be got from practical data or
mathematic models as in [1]. Put these data into a matrix and
get the antenna gain according to the elevation and the azimuth
when needed.
D. Getting Doppler Frequecy Shift
The velocity of the target can be ignored according to the
velocity of the missile. The projection of the seekers velocity

length of target is 60 meters, there is only one resolution cell.

Its return is showed in Fig. 5. Because there was no synthesis,
according to the transmitted signal, the waveform was the same
and only the amplitude changed. When the length of target is
8 600 meters, there are three resolution cells. The return is
showed in Fig. 6. Because of synthesis, according to the
transmitted signal, the waveform was widened and the
amplitude changed too.

vector onto the line from radar to target is the radial velocity
which causes the Doppler frequency shift. And the Doppler
frequency shift is
fd =2v

Echo of cell 3
Echo of cell 2
Echo of cell 1

1 2 3 . .

Q2 Q1 Q

When the target is 60 meters, choose Swerling 2 and 4

models, and do the statistic histograms of 600 pulses returns as
Fig. 7 and Fig. 9. From them we can see, the returns statistic
accorded with exponent and 2 distribution respectively.
When the target is 600 meters long, the number of resolution
cells becomes larger. Therefore the curve became more
centralized and smooth which showed some amplitudes
probability increased largely because of synthesis as showed in
Fig. 8 and Fig. 10.

resolution cell

Figure 3. Synthesis echoes of resolution cells

E. Getting the Complex Reflection Coefficient

The complex reflection coefficient of the target is = e j .
In the formula is the RCS of targets and is the phase of
target reflection characteristic. The phase of is in effect
only when there is phase vector interference between
multi-targets. It is ignored in single-target. In this thesis we
chose the Swerling 1~4 model to describe the RCS of targets as
in [2].

Figure 4. Waveform of transmitted signal

The model has two different kinds of probability distribution

and two different kinds of correlations. One is completely
correlative in the period of one scan and irrelative between
scans, as is called slow undulation. The other is irrelative
between pulses, which is called fast undulation. The four kinds
of Swerling model are exponent distribution of slow and fast
undulation model and 2 distribution of slow and fast
undulation model respectively. Once the model is chose, the
RCS can be got by inversion method.
F. Synsytheizing the Return from Minimal Resolution Cells
According to (2) and the parameters got in the steps above,
the video return from each minimal resolution cell can be got.
Because the distance from each cell to the radar is different, the
delay of each return is different. Hence, the delay of
different cell should be taken into account when synthesize the
returns. If the width of wave gate equal to the sub-pulse width
which is , the radar return of two conjoint cells has delay.
The echo of one wave gate is synthesis of the first Q 1
returns with different delay, as showed in Fig. 3.
IV.

Figure 5. Radar return from target 60m long

ANALYSIS OF SIMULATION RESULTS

Suppose the transmitted signal to be linearity frequency

modulated, the width of pulse 1 s , the bandwidth 1 MHz ,
the sampling frequency 10 MHz and the carrier frequency
1 GHz , simulating one pulse using MATLAB and the
waveform of transmitted signal is showed in Fig. 4. When the

Figure 6. Radar return from target 600m long

Figure 10. Histogram of radar return from target 600m long with model
Swerling 4

Figure 7. Histogram of radar return from target 60m long with model
Swerling 2

V.

CONCLUSION

The simulation of extended targets can describe the radar

return more exactly and really, which has important
significance in radar signal operating under some condition
where targets cannot be taken as point targets. The method can
provide more exact data for the analysis and test of radar
system.
ACKNOWLEDGMENT
Figure 8. Histogram of radar return from target 600m long with model
Swerling 2

First and foremost, we would like to show our deepest

gratitude to Dr. Li Gang, a respectable, responsible and
resourceful scholar, who has provided us with valuable
guidance in every stage of the writing of this thesis. We shall
extend my thanks to the students such as Zhang Jian and so on
in our office, who participated this study with great
cooperation.
REFERENCES
[1]

[2]
Figure 9. Histogram of radar return from target 60m long with model
Swerling 4

[3]

[4]
[5]
[6]

[7]

Wang Guo-yu, Simulation and Evaluation of Radar Electronic War

System[M](in Chinese), Beijing: National Defense Industry Publishing
Company, 2004: pp 64-212
Merrill I. Skonlnik, Radar Handbook [M], New York: McGraw-Hill,
2008: pp1-20~1-77.
Ding Lu-fei and Geng Fu-lu, Radar Elements [M] (in Chinese), Xian:
Xian Electronic Science and Technology University Publishing
Company, 2000: pp128-163.
Mitchell Richard L., Radar Signal Simulation [M], London: Artech
House, 1976.
Zheng Chong-you, Geometric Conspectus [M] (in Chinese), Beijing:
Altitude Education Publishing Company, 2005: pp 98-138
Zhang Wei, Research on Theories, Methods and Application of Radar
System Simulation [D] (in Chinese), Chengdu: Electronic Science and
Technology University Doctor Degree Thesis, 2004
Zhang Chang-long, Research on Modeling and Simulation of clutter and
Its Application in Radar Signal Simulator [D] (in Chinese), Changsha:
National University of Defense and Technology Doctor Degree Thesis,
2004