See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/265384026
Simulink Implementation of FrequencyHopping Communication System Article · December 2009
READS
475
1 author:
Kefei Liu Arizona State University 33 PUBLICATIONS 52 CITATI CITATIONS ONS
SEE PROFILE
All in-text references underlined in blue are blue are linked to publications on ResearchGate, letting you access and read them immediately.
Available from: Kefei Liu Retrieved on: 08 July 2016
© Journal of System Simulation
第 21 卷第 24 期 2009 年 12 月
Vol. 21 No. 24 Dec., 2009
Simulink Implementation of Frequency-Hopping Communication System 1
1
2
LIU Ke-fei , YANG Dong-kai , WU Jiang
(1. School of Electronics and Information Engineering, Beihang University, Beijing, 100083, China; 2. Department of System and Control Science, Beihang University, Beijing, 100083, China)
Abstract: Based on the introduction of frequency-hopping communication system and its mathematic model, a simulation model was built using Matlab/Simulink. In the simulation model, the core components of frequency-hopping system, including frequency-hopping sequence generator, frequency synthesizer, frequency hopping synchronizer and nonconherent FSK demodulator were designed and implemented. Simulation test was done with satisfied result. The performance of frequency-hopping system in various conditions could be analyzed and evaluated through the bit error rate curve of the proposed model, such as anti interferce, anti multi-path fading and multiple access networking. Key words: frequency-hopping communication; simulink simulation; frequency synthesizer; envelope detector; anti interference; multiple access networking
Simulink 刘克飞 1，杨东凯 1，吴 江 2 （1.北京航空航天大学 电子信息工程学院，北京 100083 ；2.北京航空航天大学 系统与控制科学系，北京 100083 ）
：在介绍跳频通信系统的原理和数学模型的基础上，利用 Matlab/Simulink 建立了跳频通信系统的仿真模型。
仿真结果证明了模 型的正确性。通过仿真得到的误码率曲线，可以分析评估各种条件下跳频系统的抗干扰、抗多径衰落和多址组网等 性能。 ：跳频通信系统; Simulink 仿真; 频率合成器; 包络检波器; 抗干扰; 多址组网 ：TN914.41
Introduction
：A
1004-731X (2009) 24-7969-05 military communication but also in civil mobile communication
1
such as GSM, Home RF and Bluetooth. Many factors, e.g. FH
Simulink is a platform integrated into Matlab for
sequence, coding, modulation, synchronization algorithm and
multi-domain simulation and model based design of dynamic
channel type, can influence the performance of FH system. In
systems. It provides customizable block libraries for analyzing,
scientific research, it often needs to build a simulation platform
designing, simulating, implementing and testing control, signal
to analyze the effect of a special factor on the performance of
processing, communciatios and other time-varying systems.
the FH system in terms of anti interference, anti multi-path
Simulink also provides graphic interface for modeling with
fading, multiple access networking, etc. However, no article
block diagram, allowing users to create and mask their own
has, so far, described how to build an FH system simulation
subsystems. As a visualized simulation tool, simulink is
platform in detail. In this paper, we intend to build a basic
outstanding in intuitiveness, convenience, flexibility and
simulation model of FH system with MATLAB/Simulink and
accuracy.
describe it at lenth. The model can serve as a basic platform for
With excellent anti interference, anti multi-path fading and multiple access networking performance, frequencyhopping(FH) techneque has been widely used not only in
Received date: 2008-08-25 Revised date: 2008-11-04 Foundation item: National Science Foundation of China (60602046) Biography: LIU Ke-fei (1982-), male, born in Lankao of Henan Province, Han nationality, Postgraduate, and his research interest is spread spectrum communication ； YANG Dong-kai (1972-), male, born in Laiwu of Shandong Province, Han nationality, Doctor, associate professor, and his research interest are satellite navigation signal processing algorithms, wireless data transmission methods, et al. ；WU Jiang (1982-), male, born in Baoji of Shanxi Province, Han nationality, Postgraduate, and his research interest is robot control.
analyzing and evaluating the performance of
the FH system
for various conditions.
1
Mathematic Model In the FH communication system, the transmitting
frequency is hopped
in the whole frequency band according
to certain frequency -hopping pattern. FH/FSK is the most common modulation, and it uses frequency-shift keying (FSK) modulation and non-coherent demodulation. The block diagram of FH communication system is shown in Figure 1.
• 7969 •
第 21 卷第 24 期
Vol. 21 No. 24
系
2009 年 12 月
统
仿
真
学
报
Dec., 2009
r(t) ⋅ 2cos2π f0t = w1 (t ) + w2 (t ) + n (t ) ⋅ 2cos2π f 0t Data
FSK modulator
FH modulator
FH demodulator
Frequency synthesizer
Frequency synthesizer
FH sequence
FH sequence
(4)
where
Data FSK demodulator
w1 (t ) = cos 2π ( f c + m∆f )t
w2 (t ) = cos2π ( fc + m∆f + 2 f 0 )t + U −1
∑[cos 2
( fc + m∆f + fu − f0 )t +
π
u =1
cos 2π ( fc + m∆f + fu + f 0 )t ]
If other U-1 users’ hopping carrier frequencies don’t
FH synchronizer
collide with that of the 0th user, i.e., f u ≠ f 0 (u = 1, 2,,U − 1) ,
Fig. 1 Block diagram of FH communication system
then after passing IF bandpass filter(central frequency is
Let us assume that M is the FSK modulation level, m is the M-ary information symbol data ( m =0,1,…, M -1), U is the
f c + ( M − 1)∆f /2 ), w2 (t ) is removed, and only the useful signal w1 (t ) and
noise remains, as follows
number of all active users, and T s is the symbol period. The
w(t ) = cos 2π ( fc + m ∆f )t + nc (t )
MFSK modulator output signal of the uth （u=0,1,…,U -1）user
The filtered signal passes through the nonconherent
(5)
M-FSK demodulator [3], as shown in Figure 2, and recover the
in the symbol period 0 ≤ t ≤ T s is given by xu (t ) = cos2π ( f c + m∆f )t , m ∈ {0,1,,M-1}
(1)
ˆ . information data m
where f c is the center frequency and ∆ f is frequency spacing.
⊗
In FH modulator, xu(t) is multiplied by the frequency
∫
T s
( ) dt
()
2
()
2
i
i
0
Σ
cos2πf c t
synthesizer output signal, whose frequency is generated
⊗
according to the FH sequence during a hop period, and then
2
∫
T s
T s
0
( ) dt i
i
hop period is assumed to equal the symbol period. So the FH
⊗
(2)
Decision Stage: Choose m corresponding to the maximum value
sin2πf c t
w(t)
passes through wide-band bandpass filter. For simplicity, the modulator output signal becomes su (t ) = cos 2π ( f c + m∆f + f u )t
2
T s
2
∫
T s
T s
0
( ) dt
( )2
( ) dt
2
i
i
Σ
cos2π[f c+(M-1)△f]t
⊗
where f u is the hopping carrier frequency of the u th user in the
ˆ m
2
∫
T s
T s
0
()
i
i
sin2π[f c+(M-1) △f]t
hop duration.
Fig. 2
Through AWGN channel, the received multi-user mixed signal in noise is
2
U −1
r (t ) = ∑ cos 2π ( f c + m∆ f + fu ) t + n( t)
(3)
Simulation Model Take buliding a single-user system model for example.
u =0
where n(t ) is AWGN noise with unilateral power spectral
Block diagram of nonconherent MFSK demodulator
The simulation tool is Matlab 7.1/Simulink 6.0. Parameters are shown in Table 1, and simulation model is shown in Figure 3.
density of N 0. Suppose that the 0th user is the expected user, and the local carrier is completely in synchronization. The received signal r (t ) is multiplied by the synthesizer output local carrier
To build multiple-user system model, it just needs to connect multiple single-user system in parallel and let them pass one shared AWGN channel.
signal 2cos2π f 0 t . Table 1
Simulation parameters of FH communication system
BFSK signal
System bandwidth Modulation
Date bit
Frequency interval
Number of frequency points
FH rate
Signal bandwidth
Mark frequency
Space frequency
1400Hz~11000Hz
200bps
600Hz
16
200hop/s
600Hz
600Hz
400Hz
FH/BFSK
Tx
Bernoulli Binary
BFSK Modulator
Bernoulli Binary Generator
BFSK Modulator
FH Sequence
Frequency Synthesizer
Re(u)
FH Sequence Generator
Frequency Synthesizer
Complex to Real-Imag
FH Modulator FH Modulator
AWGN AWGN Channel
FH Demodulator
BFSK Demodulator
FH Demodulator
BFSK Demodulator
2 Gain receiver
transmitter
Fig. 3
Simulink diagram of FH communication system simulation model
• 7970 •
Error Rate Calculation Rx Error Rate Calculation
Display
FH sequence m sequence
第 21 卷第 24 期
Vol. 21 No. 24
刘克飞, 等：跳频通信系统的 Simulink 仿真实现
2009 年 12 月
Dec., 2009
DSP
2.1 System Description At the transmitter, the Bernoulli Binary Generator block
Sine Wave
generates random binary data with symbol width of 1/200s. 1
Then the data feeds into the BFSK Modulator subsystem for
In1
baseband modulation. FH Sequence Generator subsystem generates
FH
sequence,
which
controls
the
1 Out1
Switch DSP
Frequency
Synthesizer subsystem to generate periodic frequency-hopping
Sine Wave1
complex exponential carrier signals. In the FH Modulator
Fig. 4 Internal diagram of BFSK Modulator subsystem
subsystem, the output complex exponential carrier signals of Table 5
The main parameters of Sine Wave and Sine Wave1 parameter name parameter value Amplitude 1 Frequency(Hz) 600 400 Phase offset(rad) 0 Sample mode Discrete Output complexity Complex Sample time 1/(600*16) Samples per frame 1
the Frequency Synthesizer subsystem and the output complex exponential signals of the BFSK Modulator subsystem are mixed together to generate a real sine wave. The frequencymixed signal is sent to the AWGN Channel. At the receiver, all users receive multi-user mixed signals in AWGN noise. The frequency hopping signals first pass the FH Demodulator subsystem for dehopping, and then pass through
PN Sequence Generator
the BFSK Demodulator subsystem for nonconherent FSK demodulation. The Error Rate Calculation block is used for
PN Sequence Generator
calculating the bit error rate, and the Display block shows the result. Parameter setting of related blocks is shown in Table 2~4. Table 2
Probability of a zero
Initial seed
Sample time
parameter value
0.5
1000
1/200
Buffer
1 Out1
Bit to Integer Converter
Unbuffer
Fig. 5 Internal diagram of FH Sequence Generator subsystem
The main parameters of Bernoulli Binary Generator
parameter name
Bit to Integer Converter
The internal diagram of the Frequency Synthesizer subsystem is shown in Figure 6. The FH sequence feeds into the Discrete-Time VCO block, whose output signal frequency
Table 3
parameter
The main parameters of AWGN Channel
Mode
name parameter
Signal to noise
value
ratio (Eb/No)
Table 4
is determined by two parameters: Quiescent frequency f and
Es/No
Input signal
Symbol
(dB)
power (watts)
period (s)
0
0.5
Input sensitivity s. To be specific, the frequency of the transient
1/(11000*1
output signal is m=f+s*u, where u is the input voltage indicated
6)
by FH sequence. The two input ports of the Variable Transport Delay block, import the output signal of the Discrete-Time
The main parameters of Error Rate Calculation
parameter
Receive
Computation
Computation
Output
VCO block and the tansmission time delay, respectively. The
name
delay
delay
mode
data
transport time delay is 1/(4*m), i.e., delay of pi/2 phase. The
2
0
Entire frame
Port
output real signal of the Discrete-Time VCO block and its
parameter value
pi/2-phase-delayed version combine to complex signal in the Real-Imag to Complex block. The complex signal feeds into
2.2 Internal Diagram of all Subsystems The internal diagram of the BFSK Modulator subsystem is shown in Figure 4. Here, the binary FSK signal is generated
the output port. Parameter setting of related blocks is shown in Table 6 and 7.
with the Digital Keying Method. The two Sine Wave blocks
Discrete-Time VCO
1
genarete complex exponential signal of frequency f 1 and f 2
In1
switches on only one of the two Sine Wave blocks, and controls
Fcn
Parameter setting of related blocks is shown in Table 5.
1 Out1
Ti Variable Transport Delay
Fig. 6 Internal diagram of Frequency Synthesizer subsystem
The internal diagram of the FH Sequence Generator subsystem is shown in Figure 5. PN Sequence Generator
Buffer block to convert into data frames of 4-bit, then passes
Real-Imag to Complex
1/(4*(f+s*u))
it to output complex exponential signal with frequency f 1 or f 2.
length of 15. The binary sequence first passes through the
Im
Discrete-Time VCO
respectively. In each symbol period, the input binary data
generates m sequence with sampling period of 1/800s and
Re
Table 6
parameter name parameter value
through the Bit to Integer Converter block to convert into data frames of hexadecimal number, and finally passes through the Unbuffer block to convert into data samples. The sample-based hexadecimal number feeds into the output port. • 7971 •
The main parameters of Discrete-Time VCO
Output amplitude(V) 1
Quiescent Input Initial frequency(Hz) sensitivity(Hz/V) phase(rad) 1200
600
0
Table 7 The main parameters of Variable Transport Delay parameter Maximum Initial Select delay type name delay Buffer size parameter Variable transport 10 1e6 value delay
第 21 卷第 24 期
Vol. 21 No. 24
系
2009 年 12 月
统
仿
真
学
报
Dec., 2009
Table 8
The internal diagram of the FH Modulator subsystem is shown in Figure 7. The product of the real part of the output
The main parameters of Digital Filter Design
parameter name
parameter value
Response Type
Bandpass
Design Method
Chebyshev type II
complex exponential signal of the BFSK Modulator subsystem and that of the Frequency Synthesizer subsystem, minus the
Filter Order
product of the imaginary part of the same two output signals,
Frequency
Minimum Order
equals the real frequency mixing signal.
Product Re(u) Re(u) 1
2
Im(u)
In1
In2
Complex to Real-Imag
Magnitude
1
Im(u)
Out1
Complex to Real-Imag1
Units
Hz
Fs
11000*16
Fstop1
175
Fpass1
200
Fpass2
800
Fstop2
825
Units
dB
Apass
0.01
Astop
80
The non-coherent BFSK Demodulator subsystem consists of mainly an envelop detector and a judging unit,
Product1
Fig. 7
whose internal diagram is shown in Figure 9. The envelop
Internal diagram of FH Modulator subsystem
The internal diagram of the FH Demodulator subsystem is shown in Figure 8. Received signal in AWGN noise is multiplied by the local carrier, and then passes through IF bandpass filter to filter out out-of-band noise and interference. For simplicity, the local carrier synchronization is realized with a direct-connected line. Parameter setting of related blocks is shown in Table 8.
detector is made up of two inphase and quadrature correlators, integrators and square-law detectors. The Discrete-Time Integrator is used to integrate the input signal on each interval [0,Ts], the Pulse Generaotr is used to reset the output state to its initial value(0), and the Transport delay block is used to make up for the transport delay. The outputs of the two Add blocks(the sample time is equal to the symbol period 1/200s) are the envelop of the input signal corrspending to the frequency f 1 and f 2 component, respectively. The judging unit
FDATool 1
compares the envelop of two signals and output bit data.
1
In1 2
Parameter setting of related blocks is shown in Table 9.
Out1
Product Digital Filter Design
In2
Table 9 The main parameters of Discrete-Time Integrator(1,2,3) parameter Gain External Initial Sample time name value reset condition parameter 2/(1/200) rising 0 1/(11000*16) value
Fig. 8 Internal diagram of FH Demodulator subsystem
K Ts Product DSP
Transport Delay
Re(u)
z-1
u
2
Discrete-Time Math Integrator Function
Im(u) Complex to Sine Wave Real-Imag
Pulse Generator
Add K Ts
Product1 Transport Delay1
1 In1
DSP
Re(u)
Product2 Transport Delay2
Im(u) Complex to Sine Wave1Real-Imag1
z-1
u
2
Discrete-Time Math Integrator1 Function1 K Ts 2 u z-1
Add1
K Ts
Delay3
Relay
Discrete-Time Math Integrator2 Function2
Pulse Generator1
Product3 Transport
1
z-1
u
2
Discrete-Time Math Integrator3 Function3
Fig. 9 Internal diagram of nonconherent BFSK Demodulator subsystem
• 7972 •
Out1
第 21 卷第 24 期
Vol. 21 No. 24
刘克飞, 等：跳频通信系统的 Simulink 仿真实现
2009 年 12 月
Fig. 10
3
Dec., 2009
Time domain waveform at every point during simulation
Simulation Result
sequence generator, frequency synthesizer, synchronizer and
3.1 Time domain waveform at each point during simulation
envelope detector. The model can serve as a basic reference model. After slight modification on it, such as changing the FH sequence, replacing AWGN block with Multipath Rayleigh,
The time domain waveform at each point during the
Rician
Fading
Channel
block,
or
adding
narrow-band
simulation process is shown in Figure 10, where E b/N0 is 0dB.
interference, the effect of these factors on the performance of
3.2 Simulation results about Bit Error Rate Performance
FH system in terms of anti-jamming, anti multi-path fading and
The simulation result about bit error rate vs. signal to
multiple access networking can be analyzed.
noise per bit curve of the FH system model is shown in Figure
Preferences:
11. The simulation time is 10s. It can be seen that the
[1]
simulation bit error rate result is reasonable.
梅文华 , 王淑波 , 邱永红 , 等. 跳频通信[M]. 北京：国防工业出 版社, 2005: 8-15 (Mei Wenhua, Wang Shubo, Qiu Yonghong, et al.
10
Frequency Hopping Communciations [M]. Beijing, China: National Industry Press, 2005: 8-15). [2]
-
e t a R r o r r E t i B
徐明远, 邵玉斌. MATLAB 仿真在通信与电子工程中的应用 [M]. 西安：西安电子科技大学出版社 , 2005: 329-333 (Xu Mingyuan,
10
Shao Yubin. Simulation Application of MATLAB in Communication and Electronic Engineering [M]. Xi’an, China: Xidian University Press, 2005: 329-333).
10[3]
Heung-Gyoon Ryu, Yingshan Li, Jin-Soo Park. Effects of Frequency Instability Caused by Phase Noise on the Performance of the Fast FH Communication System [J]. IEEE Transactions on Vehicular
10-40
-35
-30
Eb/No
-25
-20
-15
Technology (S0018-9545), 2004, 53(5): 1626-1632. [4]
Fig. 11 Simulation bit error rate vs. signal to noise per bit curve
4
王翔, 黄建国, 尹玉红. 水下跳频通信系统的建模与仿真 [J]. 系统 仿真学报, 2008, 20(2): 453-457. (WANG Xiang, HUANG JI AN-guo,
Conclusion
YIN Yu-hong. Research on the Simulation of Hopping Frequency
In this paper, we have built a basic simulation model of FH communication system with Matlab/Simulink, which realizes the core components of FH system such as FH
• 7973 •
Underwater Acoustic Communication System [J]. Journal of System Simulation (S1004-731X), 2008, 20(2): 453-457).
