Vous êtes sur la page 1sur 5



DESIGN OF A NOVEL DRFM JAMMING SYSTEM BASED


ON AFB-SFB
Jing Yang,

Xiaoxu Guo,

Yunjie Li

(School of Information and ElectronicBeijing Institute of Technology, Beijing 100081, China)

Abstract
In research of Electronic Countermeasure (ECM) recently,
Jamming source based on Digital Radio Frequency Memory
(DRFM) technology and digital channelization technology
has attracted more attention. This paper schemes out a
novel DRFM jamming system structure, which
implementing the channelized receiving with Analysis
Filter Bank (AFB) and the modulated signal reconstruction
with Synthesis Filter Bank (SFB). Jamming information
modulation is carried out in an optimized way with
improved Signal Noise Ratio (SNR) of the generated signal.
Simulation results have verified the validity of this scheme.
Keywords: Digital Radio Frequency Memory (DRFM),
Analysis Filter Bank (AFB), Synthesis Filter Bank (SFB),
deceptive jamming.

1 Introduction
Since the concept of DRFM was introduced, scholars have
made lots of efforts on the system architecture,
implementation scheme as well as performance
investigation [1]. It has found its utilization widely in fields
such as radar countermeasure and relative simulation.
Traditional DRFM jamming system has the structure
described as Figure1.


Figure1 Common structure of DRFM system
In order to intercept modern frequency-agile radar signal or
Low Probability of Intercept (LPI) radar signal, Jamming
system must have broader frequency band and higher
receiver sensitivity. Reference [2] makes use of digital
channelization technology and works out an improved
DRFM system described as figure 2.


Figure2 DRFM system based on digital channelized
receiving
In the DRFM system above, Digital channelized receiving
process is carried out on the sampled data before Signal
Detection. It can improve SNR of the detecting signal
effectively. Also the storage and jamming information
modulation on the detected signal can perform only on the
receiving channel containing the detected signal. So the
DRFM system in figure 2 features broad spectrum coverage,
good receiving sensitivity and few consumption of memory
resource.
When confront complex signal with both longer time
duration and broader bandwidth in modern EW applications,
the performance of the system in Figure 2 will be
dramatically degraded. Because it can only store and
modulate partial signal from one receiving channel when
the interested signals spectrum occupying more than one
receiving channel. It is necessary to design DRFM system
which can not only maintain the advantage of channelized
receiving, but also can deal with possible broad bandwidth
signal occupying multiple channels.
Analysis-Synthesis Filter Bank (ASFB) is an important
branch of multiple-rate digital signal processing theory, and
has found its applications in many fields as references [3-6]
mentioned. AFB process input signal with a group of
filtering channels followed by corresponding decimation in
serial, while SFB can implement signal reconstruction with
a group of interpolators and corresponding filtering
channels. They may be used to design one novel DRFM
jamming system. Combining ASFB implementation with an
optimized channel modulation method, this paper
introduces one novel DRFM jamming system design
method. Applicable structures of AFB and SFB under
over-sampled condition are derived in Section 2. Jamming
system structure and deceptive information modulation

special for radar ECM applications are introduced in


Section3. Section 4 put forward some simulations of the
designed DRFM function and jamming information
modulation.

Denoting the content in the curly brackets from Equation

(5) as:
nS

P 1

j
H l ( w ) = h ( n K  l )e  j w n K e 2

(6)

n 0

2 Analysis-Synthesis Filter Bank Architecture


for NPR
ASFB is the most important part in the designed DRFM
jamming system.
Based on the assumption of input signal represented in
complex format, reference [6] brought out an efficient
implementation way for digital channelized receiver with
K channels in Fast Fourier Transform (FFT) format. But as
to the applications in real world, they always have real
signal as their input. On this occasion, half of the total
divided channels in the structure from [6] will have
redundant information. Reference [7] then proposed a
channelized receiver architecture for real input. According
to its channel division, every channel has useful signal and
the central frequency of every channel is:
2K  1 2S
k ( k 
) , k 0,1...K  1
(1)
4
K
Channel number K equals to decimate index M in the
architecture defined in reference [7], which means the
corresponding Discrete Fourier Transform (DFT)
filter-bank is implemented under a critical-sampled
condition. Over-sampled DFT filter-bank with K>M has
better performance over the critical-sampled one in Near
Perfect Reconstruction (NPR) applications. Therefore, this
paper derives the DFT filter-bank structure under
over-sampled condition for real signal firstly.
In digital channelized receiver, the input signal will be
] k 0,1, 2...K  1) in K different
filtered by h k [ n

Then
K 1

Yk ( w )= { X ( w ) e  jwl }e

K 1

 jl
y k ( n )= {{ x ( n  l ) hl ( n )}e

X(z)

2 K 1 2 S
4 K

}e

H0(zF )

k 0

n 0

H1(zF )

k 1

n 1

Z-1
......

......

......

......

......

......

......

HK1(zF )

IFFT

 jnwK

(2)

k 0

k K1 n K1

n 0

H0(zF)

P is the ratio of the length of prototype filter h (n) to the


channel number K.When transfer function of hk [ n] is:

H k (w) H (w  wk ), k 0...K 1

(4)

Yk (w)=X(w)e e
l 0

 jwl

2Skl
2Sl (2K1)
j
K
4K

nS
j
P1 
 jwnK
h(nK  l)e e 2
n 0

yK1(n)

n 1

H1(z )

k 1
FFT

......

......

......

......

M
......

Z-1

......
Z-1

HK1(zF)

X (z)

Figure 4 Synthesis Filter Bank structure

3 Novel DRFM Jamming System and Its


Jamming Modulation
Based on ASFB implementation from Section 2, one novel

Then, output of every branch filter is:


K1

y1(n)

Z-1

......

(3)

n 0

y0(n)

Figure 3 Analysis Filter Bank structure

yK1(n)
k K1 n K1

h ( nK  l )e

(8)

......

Z-1

Where
P 1

2S
K

......

l 0

El (e jwK )

jkl

Z-1

......

El (e jwK )

(7)

Considering k and l as variables, Equation (8) is


equivalent to an Inverse Fast Fourier Transform (IFFT)
operation. When K is even, basic diagram of digital
channelized receiver can have its realization in poly-phase
structure in which the decimation operation has been
carried out just before poly-phase filter in every channel
according to Reference [8]. Diagram of AFB realization
structure shown in Figure3 can be acquired. SFB has a
coupled architecture to AFB, which is described in Figure
4.

y1(n)

 jwl

2 S kl
K

l 0

Given the used prototype filter as h[ n] , its equivalent


poly-phase representation is:

j
H l ( w ) e

The equivalent expression in time domain is:

y0(n)

H ( w)

2 S l ( 2 K 1)
4K

l 0

processing sub-channel. All branch filter hk [ n] extract


from one same prototype filter h[ n ] in a poly-phase way.

K 1

j

DRFM jamming system can been schemed out. Figure 5

(5)

shows its main diagram.

Radar Signal
Parameters
Measurement

Delay Control
for Channel A
N

r fd A _ i
JA-buffer-1

Channel
A

       Figure 5 Realization structure in serial


Deceptive jamming applications usually need to add some
false information intentionally onto the interceptive signal.
Time delay in range dimension and Doppler shifting in
frequency dimension are two basic kind of information to
be modulated.
There is no variable associated with time in our AFB-SFB
architecture, so delaying the output of AFB for certain time
period before sending it into SFB will not put any effect on
the signal reconstruction. For jamming system realization
based on DRFM structure, reading the intercepted signal
out from memory at exact time can fulfill the deceptive
information modulation of distance.
Doppler frequency modulation relies on the realization of
AFB and SFB. AFB analyzes input signal in interested
whole spectrum range and SFB reconstructs the signal
based on the output from AFB. If there is frequency shifting
f
of d Hz in every channel, the reconstructed signal will
f
also have d Hz frequency compared with input signal.
Spectrum shifting in frequency domain is equivalent to
multiplication operation in time domain, so proper multiply
operation at the output of every channel can fulfill the
Doppler modulation implementation.
The frequency modulation method proposed above may
result in some amplitude distortions at the overlapped
boundary between neighbor channels. Analysis and
simulation have shown that only when the offset of
frequency shifting is large enough to the Megahertz level,
the possible distortion can lead to serious impact on the
system performance. Doppler modulation in common
jamming applications always has the frequency shifting no
more than hundreds of KHz in usual jamming system
applications, so this method is applicable.
Connecting AFB, SFB and the Jamming Information
Modulation module together, we can establish the
fundamental part of the designed jamming system.
Supposing the input signal has the bandwidth occupying
two continuous channels at most while modulation of N
deceptive targets in distance and two deceptive targets in
Doppler frequency domain are wanted, Figure 6 and Figure
7 give out two possible implementation schemes. Figure 6
generates multiple deceptive jamming targets in range with
block memory buffers working in parallel and Figure 7
generate
these
jamming
signal
through
some
First-In-First-Out(FIFO) memories working in serial. Both
of the two modulation architecture can realize deceptive
Doppler modulation in frequency domain.

Data
to
ADC

Analysis
Filter Bank

Signal
Detecting
&
Adaptive
Channel
Selecting

i 1
N

r fd B _ i
JB-buffer-1

i 1

0
N

r fd A _ i
JA-buffer-2

Channel
B

i 1

Synthesis
Filter
Bank

Data
to
DAC

0
N

r fd B _ i
JB-buffer-2

i 1

Delay Control
for Channel B

Figure 6 Realization structure of modulation in parallel

Radar Signal
Parameters
Measurement

Delay Control

Channel
A

Data
to
ADC

Analysis
Filter Bank

Signal
Detecting
&
Adaptive
Channel
Selecting Channel
B

Buffer-1
N

r fd A _i
i 1

FIFO

0
N

r fd B _ i

i 1

+
Buffer-2

FIFO

Synthesis
Filter Bank

Data
to
DAC

0
N

r fd A _ i
i 1

r fd B _ i
i 1

Figure 7 Realization structure of modulation in serial

4 Simulation and Performance Evaluation


4.1 ASFB DRFM function verification
The simulation prototype is established as Figure8. Only the
channels enclosing the input signal have the connecting
paths between AFB and SFB, and the others input port in
SFB are fed with zeros directly.

x n

Figure 8 Verifying simulation of synthesis with


partial channel
Taking a Chirp pulse train for simulation, it has pulse width
of 10 P s and modulated frequency band from 100 MHz to
300 MHz. The spectrum of the input Chirp signal will
appear in a group of channels as [5,6,7,9,10,11,12]. A

comparison is made between the input Chirp signal at AFB


input port and the x(n) output from SFB and it shows that
the method can deal with the broadband signal.

74
72
70

68

Amplitude, dB

dB

-50

66
all channels synthesis
single channel synthesis

64

-100

62

-150
60

-200

-250

58
50

Spectrum of filter bank input


Spectrum of filter bank output
0

200

400
600
800
Frequency, MHz

1000

1200

Figure 9 Result of partial channel synthesis for


broadband Chirp signal

100

150

200

250

300
MHz

350

400

450

500

550

Figure 11 SNR improvement in partial channel synthesis


The result shows that the operation of adaptive synthesis
channels selection leads to SNR improvement by 12 dB in
the synthesized signal, which is accordance with the fact
that noise bandwidth is reduced by 1 / 16 .

4.2 Deceptive Doppler frequency modulation


Input signal x( n) is a pulse train. Its carrier frequency is
168.75 MHz, pulse width is 2

Ps

and pulse repeat period

is 8 P s . Doppler frequency shifting of 30 KHz is made in

the modulation module. Doppler Frequency Detection


shows the Doppler modulation proposed here is applicable.
0
X: 30
Y: 0

Amplitude, dB

-20

-40

-60

DRFM has been used widely in radar countermeasure and


signal simulation fields. This paper puts forward one novel
DRFM Jamming System design scheme. On the basis of
AFB and SFB implementation design under over-sampled
condition of K>M, this scheme employs AFB for
channelized receiving of input signal and SFB for generated
jamming signal reconstruction. At the same time, one
simplified modulation step between AFB and SFB can
perform the jamming signal generation properly with higher
SNR. Simulation and practice have proved that the method
proposed is effective and applicable.

References

-80

-100

5 Conclusions

20

40
60
Frequency, KHz

80

100

Figure 10 Detecting result of Doppler frequency


modulated signal

4.3 SNR improvement in partial channels synthesis


For an AFB and SFB implementation with 16 channels,
simulation has CW signal contaminated by noise as the
input signal. The frequency of the input signal changes
from channel 1 to channel 16. Under the condition of input
signal with SNR as 60dB, results comparison between total
channel synthesis and partial channel synthesis are plotted
together in Figure12.

[1] S. J. Roome, "Digital radio frequency memory,"


Electronics & Communication Engineering Journal, vol. 2,
pp. 147-153, 1990.
[2] Sun Guoying, Li Yunjie, Gao Meiguo. An improved
DRFM system based on digital channelized receiver[C].
Congress on Image and Signal Processing 2009:
3617-3620.
[3] P. P. Vaidyanathan, Multirate systems and filter banks:
Prentice-Hall, Inc. Upper Saddle River, NJ, USA, 1993.
[4] L. Xiaojian, "Acoustic echo cancellation over nonlinear
channels." vol. PH.D: McGill University (Canada). , 2004.
[5] P. P. V. Yuan-Pei Lin, "Application of DFT filter banks
and cosine modulated filter banks in filtering," 1994.
[6] Y. Shi, "Design and Estimate for Multirate Systems."
vol. PH.D: University of Alberta, 2006.
[7] Zahirniak D R, Sharpin D L, Fields T W. A
Hardware-Efficient, MuItirate, Digital Channe1ized
Receiver Architecture; IEEE Transactions on Aerospace
and Electronic Systems, VOL.34, NO.1, January 1998.

[8] Xu Hua, Lu Wusheng Efficient Iterative Design


Method for Cosine-Modulated QMF Banks; IEEE
Transactions on Signal Processing, VOL.44, NO.7, July
1996.

[9] Qi. Zhang, Research and Realization of Broadband


Digital Receiver, Beijing Institute of Technology, 2006.
[10] Zong Bo. Wang, Research on Jamming Techniques
for Broadband and Widepulse Signal, Beijing Institute of
Technology, 2009.

Vous aimerez peut-être aussi