Communications and Network, 2013, 5, 15-21
http://dx.doi.org/10.4236/cn.2013.53B2004 Published Online September 2013 (http://www.scirp.org/journal/cn)
A Simulation Study on Channel Estimation for Cooperative
Communication System in Sand-dust Storm Environment
Xuehong Sun1,2, Yu Cao1, Jin Che1
1School of Physics and Electrical Information Engineering, Ningxia University, YinChuan, China
2School of Information and Communication Engineering, Beijing University of Posts and Telecommunications, Beijing, China
Email: nxsunxh@gmail.com
Received May, 2013
ABSTRACT
There are many factors that influence the propagation of electromagnetic wave in the sand-dust storm environment, the
scattering effect of dust particle is one of the major factors, so this paper focuses on the dust particles scattering function.
The scattering of dust particles inevitably brings the multipath transmission of the signal, multipath propagation will
bring the multipath fading of the signal. In this paper, we first investigate the use of AF and DF modes in a sand-dust
storm environment. Secondly, we present a low-order modulation method should be used in cooperative communication
system. Lastly, we evaluate the system performance for both of the moving nodes and power allocation. Experimental
results validate the conclusion of theoretical derivation: the multipath fading is one of the main factors that affect the
quality of signal transmission. Cooperative communication technology has good anti-fading ability, which can guaran-
tee the signal transmission timely and correctly.
Keywords: Cooperative Communication; Sand-dus t Storm; Particle Scattering; Multipath Fading; OFDM
1. Introduction
During the wireless communications, the main factors
that influenced the quality of wireless commun ication are
the complexity of the transmission environment and mul-
tipath fading of wireless signals. The transmission envi-
ronment in the sand-du s t storm is more complex than that
in the urban environment. There are many factors that
can influence the transmission of electromagnetic wave,
firstly, the loss of signal transmission in the sandstorm
channels; secondly, the attenuation that caused by the
water content and visibility of the sandstorm; thirdly, the
polarization effect of dust particles influence on electro-
magnetic waves. Last, the scattering effect of dust particles.
This paper discusses cooperative communication sys-
tem in sand-dust storm environment. It does research on
sand-dust storm communication system, for the project
of Ningxia University The research on access to infor-
mation technology cooperation in desert hinterland dust
monitoring.
It is a project for information acquisition of desert and
sand-dust storm. The implementation environment of the
project is in desert hinterland, about 25 kilometers. It
uses the sensors carried by robots to get a variety of data
in desert. The data is transmitted to remote information
center by wireless communication network. In desert
hinterland, each node can be real-time mobile robots.
With the moving of the nodes, the communication net-
work is also changing. When a node is moving, a node
must leave or enter the coverage situation of other com-
munication nodes, information transmission is required
in the cooperation of other nodes to complete. In addition,
the nodes may move into the position of the sand dunes,
which will cause that the information cannot reach the
destination node by direct passing way, this also need
other nodes together to complete the data transmission.
When the electromagnetic wave is transmitting in the
sand-dust storm environment, it will be affected by dust
particles. Dust particles will cause the scattering effect.
The scattering of dust particles inevitably brings the mul-
tipath transmission of the signal, multipath propagation
will bring the multipath fading of th e signal.
In this paper, we do the study of cooperative commu-
nication mode and dust environment, and look for a
suitable cooperative communication system, which can
work in sand-dust storm environment.
2. Theory Explanation
2.1. Cooperative Communication Model
2.1.1. AF model and its performance
Amplify-and-forward (AF) is one of the most simple way
in cooperative communication. It is same as Decode-
and-forward (DF) model in the first time slot. However,
C
opyright © 2013 SciRes. CN
X. H. SUN ET AL.
16
in the second time slot, it amplifies the signals received
by relay and sent it directly. And use MRC (Maximum
Ratio Combining) to do signal processing for the first
and second time slot[2 ].
1, 2,
s
d
yay ay
rd
(1)
where
,,
22
,0
,
12
22
0,2
0
22
,0
,
1
sr
s
rrd
sr s
ssd
rrd
sr s
PP hh
hP
Ph
aa
Ph
hP






marks conjugate transpose, the instantaneous SNR is
12

 [1], 1
is the SNR of the source node to
destination node direct passing channel, 2
is the SNR
of relay node to destination node.
2
1
2
0
ssd
Ph
,
(2)
22
,,
22
00
222
,,
22
00
1
sr
sd rd
sr
sd rd
PP
hh
PP
hh


(3)
According to the reference [2], instan taneous 2
is:
22
,,
22
00
222
,,
22
00
ˆ
sr
sd rd
sr
sd rd
PP
hh
PP
hh


(4)
And use MQAM (Multiple Quadrature Amplitude Mod-
ulation) to do the analysis of the AF mode.
MQAM is a kind of carrier wave control mode that
widely used in the cap acity and the large capacity digital
microwave communication system. It can improve the
spectrum efficiency of communication system. At pre-
sent, in large capacity digital microwave communication
system, such as SDH (Synchronous Digital Hierarchy)
digital microwave and LMDS, there are widely used in
64 QAM and 128 QAM, which are MQAM modulation
mode.
According to the reference [3], the symbol error rate
(SER):




,,,12
22 12
ˆ
4
ˆ
4
sd sr rd
hhh
MQAM MQAM
MQAM
PKQb
KQ b




(5)
where
13
1, 1
MQAM
Kb
M
M

Q is

2
2
2
0
1exp 2sin
z
Qz d





2
4
22
0
1exp 2sin
z
Qz d




Average SER is :





22
22
20, ,
0
0,,,
22
22
240, ,
0
0,,
4
0, ,
22,,
4
02
2sin
2sin
4
2sin
2sin
4
2
1
2
AF
MQAM
ssr rrd
MQAMssdMQAMsrsrr d
ssr rrd
MQAMssdMQAMsrsrr d
ssr rrd
MQAMsrsrr d
MQAM
P
PP
Kd
bP bPP
PP
K
bP bPP
AP P
bPP
A
b


,











,, ,
11
ssdssrrrd
PP P



 

(6)
where
22
24
44
00
31
44
sinsin 8
M
KK
Ad d
M

 
K

 

(7
The Figure 1 is the performance curve of 16 QAM
4
)
and
QAM modulation system in AF mode, where ,sd
,,
1
sr rd
 , 2
01
, 34K, 1615
QAM
b. Cur
e pea4Qter than
16QAM. In low SER, the difference of two methods is
not bit. With the increase of SNR, the good performance
of 4QAM gradually reflected. Therefore, the low-order
modulation method should be used in AF mode.
ve
shows that thrformnce of AM is bet
,
Figure 1. The performance curve of 16 QAM and 4 QAM
modulation system in AF mode.
Copyright © 2013 SciRes. CN
X. H. SUN ET AL. 17
2.1.2. DF Model and Its Performance
Decode-and-forward(DF) is another way in cooperative
communication. In DF model, the relay receives the sig-
nal and does demodulation and decoding. Then, the ver-
dict will be done; the relay will do recoding for the data
and send it out again. The study of DF model also can
use MRC.
1, 2,
s
dr
yay ay
d
where
 (8)
2
1,0ssd
aPh
, 2
2,rrd
aPh0
The SNR of the receiver is:
2
,ssd
Ph
2
,
2
0
rrd
Ph
(9)
In MQAM modulation, the approxim
system is: ate SER of the
,
22
22 22
00
2,,
244
0
,
222
0
,
222
0
2sin 2sin
4sin
112sin
2sin
MQAM ssdMQAM ssr,
MQAMsrs drd
MQAMssr
MQAMss d
FF
bPP
F
bP
F
bP
F
 













 





DF
MQAM
bP bP
P

,
222
0
2,,
244
0
24 4
00
22 2
,, ,,
2sin
4sin
44
MQAMssr
MQAMsrsdr d
M
QAM ssdsrMQAM srsdrd
bP
F
bPP
F
BC
bP bPP












 
(10)
where
2
22
2
04
2
44
sin sin
1
2
KK
Bd
M
MK
M

2
d




(11)

2
22
44
04
2
44
sin sin
31
8
KK
Cd
M
MK
M

d




(12)
The Figure 2 is the performance curve of 16 QAM
and 4 Qation systems in DF mode. It is the
sa AM modul
me as AF model.
Where ,, 1
sd s
, 21
,
,r rd034K
,
16 15.
QAM
b
Figure 2. The performance curve of 16 QAM and 4 QAM
modulation system in DF mode.
efore, the low-order mod-
lation method should be used in DF mode, as well.
the
SE
magnetic
e scattering and ab-
n of
electroic wave. Because the shapes of the dust
ding to the equivalent medium theory, dust par-
ticles can be equivalent to homogeneou medium. It is
assumed that the radius of all particle are the same, ac-
co
With the increase of SNR, the good performance of 4
QAM gradually reflected. Ther
uIn the research of AF and DF system performance, it
can draw the conclusion: in the case of high SNR, the
low-order modulation mode should be used to keep
R of the system in the low condition.
2.2. Influence of Propagation
2.2.1. Polarization Influence on Electro
Wave
From the references [4] and [5], th
sorption of dust particles can cause the attenuatio
magnet
particles are irregular, when the electromagnetic wave
spreads in the sand-dust storm, differential attenuation
and differential phase shift is on the different directions.
Dust particles can be seen as a spheroid, the ratio of its
axis is:

2
::1:0.75: 0.75abc
Accor s
rding to the reference [6], we can get the following
formula:
1
0.6287 Im m
e
fa

 
(13)
2
bm
V

1
4.15 Re 2
m
e
bm
fa
V

where
(14)
m
is the dielectric constant of dust particles, e
a
Copyright © 2013 SciRes. CN
X. H. SUN ET AL.
18
is the equivalent radius of dust particles, is the visi-
bility of sand-dust storm, b
V
is the at
cient, tenuation coeffi-
is the phase shift coefficient.


max 3
min
m2
min
a
a
ea
a
apada
aapada
(15)
ax
on
d
are:
In sa-dust storm envirment, the differential at-
tenuation coefficient and the ifferentia
of electromagnetic wave
nd l phase shift [7]

2
2.099 10
1
hv e
b
aa a
V
LL
2
sin
12 3
2
f
b
La
 
 
(16)






3
3
1.3848 10
1
2
hv e
b
La

12
LL sin
b
fa
V
 



(17)
where

 
the is the incident aof electromgnetic wave,
is equivalent radi dust particles (
visibility of dust storm (Km), f is theequeof
the incident electromagnetic wave (GHz).
ngle
us of
s
a
fr
L
e
a
the m), b
V is
ncy
and L
i i
[8] are:



1
Re 11
i
im
LA





1
im

0.324
the relationship
ent and
Im 2,3
11
i
Li
A


 

(18)
where
The ion in the
sand-dust storm environmvisibility.
1,
0.432
attenuat
1
A0.243,
Figure 3 is 2
A, 3
A.
of 3
(
v
mkgm )
is the water content in the sibility of
orms. Tcurve swhe r the visi-
bi rse.
at
t particles ine air, whiccan cause the
scattering of electromagnetic wave, it reshoots in the
dust particles arg
dust,
ho
th
0b
V is the vi
high
h
e irregular, the scatterin
the dushe s that tet st
the shapes of
lity is, the smaller the attenuation rate is; with the in-
crease of water content, the attenuation is get ting wo
According to the curve in the Figure 3, it can be seen
that in the dry dust storm, the influence that the visibility
affects on attenuation is not very clear. However, when
the water is in the dust storms, it will seriously affect the
tenuation of electromagnetic wave. In the desert hin-
terland, it can be thought that there is no water in sand-
dust storm.
2.2.2. Sc a tt er ing Influence on Electromagnetic Wave
When the sand-dust storm happens, there are a large
number of dus
attenuation of electromagnetic wave, as well. Although
statistics of the dust particle can be equivalent into a
sphere particle scattering. So, the analysis of the dust
particles can take advantage of the Mie theory [9] for
analysis. According to the Mie theory, the scattering and
extinction of section are:


222
121
2
snn
n
nab

(19)


2
121
2
t
n
Re
nn
nab (20)
where
b
In the sand-d
is the wavelength of electromag
and are the scattering coefficient of Mie.
ust storm, the distribution of dust particles
present in certain scale, the attenuation o
quency caused by unit distance (characteristic attenuation)
netic wave,
n
an
f carrier fre-
is:
r

2
1
30
4.34310( )
t
r
A
Nrrdr


(dB/km) (21)
where
r
is the size distribution probability density
function dust particles, N is the density of dust parti-
cles 0
. pevolume [10]
Because the measurement of
general study, it often uses visibility to describe the con-
on of
r unit
0
N is difficult, so in
centrati dust particles,
0
15
b
V
(22)

2
1
32
00
8.686 10r
r
Nrrdr

 (23)
where 1
r, 2
r are respective
ly minimum radius and the
maximof the dust um radius particles, 0
is the at-
agnetic wave [11]. tenuatioicient of electrom
According to (22) and (23), dust parti
unit volume ) can be calculated:
n coeffcles count per
(0
N
Figure 3. The relationship of attenuation in the sand-dust
storm environment and visibility.
Copyright © 2013 SciRes. CN
X. H. SUN ET AL. 19

2
1
3
032
22
,,
15 num m
8.686 10
[, ][10, 1].
r
br
sr rd
NVrrdr


()
(24)
The attenuation of electromagnetic wave scattering
characteristics in the sandstorm environment as follows:
 

0
2
0
15 (dB/km)
2
D
t
D
b
rrdr
AVrrdr


(25)
It can be used single dust particles attenuation for the
overall calculation [12]. The spread of the electromag-
netic wave will be influenced by multiple scattering; it
can make use of Markov process. Because electromag-
netic wave has the characteristic of light, the sample
space after scattering m times space spread sequence is
(26)
is the probability that electromagnetic
propagation in the sandstorm environment after scatter-
ing m times [13]. Because this process can be seen that
ra


1,
l
sl…,m.


0m
m
PsPs

m
Ps wave
ndom Markov process


0101mm
PsPsPs sPss
(27)
The conditional probability

11, ,
ll
Ps slm
gnetic wave prop
is
one probability of electroma
space to space after scattering, the estimation function is:
agation in

00
1
exp cos
cos
m
tm mt
mm m
m
mll
l
hz
PPWC
hz z
 



 




(28)
where tt
CN
 
10
0
x
xx

(29)
0
The weight function m
W is:
1
1exp mm
mm
zz
WW C
cos m

where
(30)
m
is a angel that scattering direction
ax scattering m times [1
and the z
is after electromagnetic wave4].
The initial weights of Electromagnetic wave is 01W
.
CN
The average transmittance of electromagnetic wave is:
1t
TP
N
(31)
Under the condition of the atmosphere visibility is
higher, the amount of dust particles in the air is small,
and the scattering of electromagnetic waves is very weak,
scattering need not to be considered. However, w
sand-dust storm is very heavy, the air visibility is less
than 1 km, the scattering of dust particles is very strong,
iously affecthe signal transmission.
3. Simulation Results and Analysis
The location of the relay node will affect the perform-
anee n of the relay node are
anFirst: the relay node is located at the source nod
destination node precisely symmetric center
That is
hen the
it will sert
ce of the system. Thrlocatio
alyzed [15]. e and
position.
22
,,
[, ][1, 1].
sr rd
The seco nd: the location of the relay node is
the close distance to destination node. That is located in
.
22
,,
[, ][1,0]
sr rd 1
The third: the location of the relay node is located in
the relatively close distance from the source node. That is
, 1].
22
,,
[, ][10
sr rd
The Figure 4 is the performance curve of relay located
in different position. It can be obtained from the figure: a
relay node is located in the close distance between source
node and destination node position, cooperative commu-
nication system performance is better, and in both cases,
the performance of system is approximately the same.
When the relay nodes in the source node and destination
node center position, performance of the system is worse.
It is the cause of the equal power allocation.
The Figure 4 is the performance curve in the equal
power allocation. The Figure 5 is the performance curve
in the unequal power allocation. The power allocation is:
the source node power is two-thirds of the total power,
Figure 4. The performance curve of relay located in differ-
ent position and equal power allocation.
Copyright © 2013 SciRes. CN
X. H. SUN ET AL.
20
Figure 5. The performance curve of relay located in differ-
ent position and unequal power allocation.
the relay node power is a third of the total power. It is
discussed in accordance with the above three kinds of
relay location conditions.
We can obtain from the Figure 5: in ranging from
he same as close
distance from destination node. Under the condition of
the same SNR, close distance relay node to destination
node position can bring good system performance. When
the relay node is in the close distance to destination node
position, collaborative system of communication system
performance is best.
4. Conclusions
In this paper, we have analyzed two kinds of cooperative
communication modes and get the conclusion that the
low-order modulation method should be used in coopera-
tive communication. In the analysis of dust environm
polarization and scattering are the two main factors t
influence the propagation of electromagnetic wave. The
moving of nodes and power allocation will affect the
lectromagnetic
nvironment is very complex, dust particle scattering of
ultipath
fa
rted by the Nature Science Found a-
62020). The authors would like to
rees for their valuable com-
power allocation, the system performance of close dis-
tance from the source node is no longer t
ent,
hat
performance of collaborative communication system.
In sand-dust storm environment, the e
e
electromagnetic waves become the important factor to
influence the signal transmission. When the signal is
transmitting, the scattering effect can cause the m
ding. So the key point of the sand-storm communica-
tion is to overcome the influence of multipath fading.
Furthermore, in the future study, OFDM technology
can be used in system, because OFDM technology can
achieve the role of resistance to multipath fading. OFDM
cooperative communication technology can support the
quality of communication system.
5. Acknowledgements
This work was suppo
tion of China (611
thank the anonymous refe
ments and suggestions.
REFERENCES
[1] A. Goldmith, H. W. Yang, W. D. Li and W. B. Guo,
“Wireless Communication,” People's Posts and Tele-
communications Press, 2007.
[2] J. N. Laneman, D. N. C. Tse and G. W. Wornell. “Coop-
erative Diversity in Wireless Networks: Efficient Proto-
cols and Outage Behavior,” IEEE Transactions on Infor-
mation Theory, Vol. 50, 2004, pp. 3062-3080.
doi:10.1109/TIT.2004.838089
[3] M. K. Simon and M. S. Alouini, “Digital Communication
over Generalized Fading Channels: A Unified Approach
to Performance Analysis,” IEEE 14th International Con-
ference on Communication Technology, Vol. 86, 1998,
No. 9, pp. 1860-1877.
isturbing Radio Wave’s
e desert, Vol. 18, No. 4, 1998,
Links,” Journal of China Institute
od Communications, Vol. 12, No. 51991.
[8] T. S. Chu, “E Microwave Prop-
[4] L. Hao, W. Yang and T. W. Wu, “The Opt ical Properties
of Dust Aerosols and Radiation Force Effect,” The Chi-
nese desert, Vol. 30, No. 6, 2010, pp. 1477-1782.
[5] N. Huang, X. J. Zheng and G. T. Chen, “Advances on
Research of Sand-dust Storm D
Propagation,” The Chines
pp. 755-353.
[6] X. M. Wu, “Sandstorm Affect Microwave, Millimeter
Wave Propagation,” Xi 'an University of Electronic Sci-
ence and Technology, 1993.
[7] W. Y. Yin and J. M. Xiao, “The Effects of Sand and Dust
Storm on Microwave
ffects of Sand Storms on
agation,” The Bell System Technical, Vol. 58, No. 2, 1979,
pp. 549-553. doi:10.1002/j.1538-7305.1979.tb02234.x
[9] R. K. Yang, D.-J. Jian and R.-H. Yao, “Research on At-
of Hygdroscopic Water on Dielec-
tenuation and the Two-Frequency Mutual Coherence
Function for Millimeter Wave Propagation in the Sand
and Dust Storm,” Journal of Xi 'an University of Elec-
tronic Science and Technology (Natural Science Edition),
Vol. 12, No. 34, 2007, pp. 953-957.
[10] J. W. Craig, “A New, Simple and Exact Result for Calcu-
lating the Probability of Error for Two-dimensional Sig-
nal Constellations,” IEEE MILCOM, Boston, MA, Vol. 25,
No. 5, 1991, pp. 1-5.
[11] S. I. Ghobrial, “Effect
tric Constant of Dust at X-band,” Electronics Letters, Vol.
16, No. 10, 1980, pp. 393-394.doi:10.1049/el:19800277
[12] X. J. Zheng, Z. W. Gao and Y. H. Zhou, “Electromag-
. C. Li and L. Xie, “Cross Depolarization
netic Wave Monte Carlo Simulation of Propagation in a
Sandstorm,”Journal of science in China, Vol. 38, 2008,
pp. 955-961.
[13] X. J. Zheng, X
Effect of Electromagnetic Wave by Partially Charged
Copyright © 2013 SciRes. CN
X. H. SUN ET AL.
Copyright © 2013 SciRes. CN
21
-570.
et. Conference
-
Spherical Sands in Sandstorms,” Journal of Desert Re-
search, Vol. 31, 2011, pp. 567
[14] G. W. Wornell and J. N. Laneman, “Energy-efficient
Antenna Sharing and Relaying for Wireless Networks,”
In Proc. IEEE Wireless Comm. And N
(WCNC2000). Chicago, USA, 7-12 September 2000.
[15] F. Elis a. “A New Permittivity Model for the Microwave
Moisture Measurement of Wet Sand,” 33rd Europe a Mi
crow Ave Conference-munich, 2003, pp. 539-542.