INTERNATIONAL CONFERENCE ON “CONTROL, AUTOMATION, COMMUNICATION AND ENERGY CONSERVATION -2009, 4th-6th June 2009

Multipath Interference Mitigation Technique for

MC DS/CDMA Systems
1
Prabaagarane Nagaradjane, 2Arun Swaminathan, 3Dhyaneshwaran.K.S, 4Boopathi R Narayanasamy and
5
Arvind Ramakrishnan

Abstract— Today in this information avaricious society, internet traffic such as Multimedia streaming application is driving the demand for
high speed data packet wireless services. In order to meet this demand and ameliorate the capacity of future generation wireless cellular
network, several capacity improvement techniques that mitigate interference have been proposed in the recent past. The capacity of MC
DS/CDMA system is limited by multipath fading and Multiple Access Interference. Mitigating MAI can improve the overall capacity of the sys-
tem. In this paper, a signal modeling for unknown code channels using OVSF code characteristics is analyzed with a Multipath Interference
Mitigation scheme that can exterminate the unknown intracell interferences additionally based on this signal modeling. The main objective of
the work is to mitigate MAI in such a way that the orthogonality is preserved to the maximum extent and elevate Pedestrian mobility needs. It
is demonstrated that the proposed scheme achieves significant performance improvement for increased number of users in accession to
maintaining orthogonality.

Index Terms— MPIM, MC DS/DMA, OVSF, UMTS

—————————— ‹ ——————————

1 INTRODUCTION

T HE future generation wireless systems that are

to be rolled out aims at offering greater capacity,
higher data rates and a wider mix of communi-
scheme, Section 4 expounds the the performance anal-
sis and section 5 draws the conclusion.

cations services compared to the second generation

systems. MC DS/CDMA has been considered as the
2 SYSTEM MODEL
mainstream air interface solution for the 4G networks.
The new characteristics and the flexibility to introduce
new services offered by MC DS/CDMA systems are to
be exploited by a variety of mobile devices and innova-
tive seamless applications. However in MC DS/CDMA
the high data rate users may cause large Multiple
Access Interference (MAI) due to their increased
transmission power [1]. This problem is particularly
severe in the Uplink (mobile to Base station) where the
users transmit asynchronously, producing the well-
known near-far problem. In addition to MAI, the re-
ceived signal is also subjected to Multipath fading due
to the relative motion between the mobile and local
scatterers. Also, distance dependent path loss and long
term fading (shadowing) exist. All of these significantly
degrade the Uplink system performance. A novel Mul- Fig.1. MC-DS/CDMA Transmitter.
tipath Interference Mitigation (MPIM) scheme that is
robust is analyzed in this paper. The MPIM scheme in 2.1 Mathematical Model
accession to turbo coding is shown to have perfor-
mance very close to Shannon's limit as the number of Figure 1 shows the generation of a multi-carrier direct
iterations increases, thereby bringing down the error sequence spread spectrum signal. The data symbol rate is
rate in the presence of MAI[2 3 4]. 1/Td. A sequence of Nc complex-valued data symbols
d n , n = 0. . . Nc − 1, of user k is serial-to-parallel converted
(k )

The paper is organised as follows. Section 2 introduces
the system model, Section 3 gives the gist of MPIM-
————————————————
onto Nc sub-streams. The data symbol rate on each sub-
stream becomes 1/(NcTd ). Within a single sub-stream, a
data symbol is spread with the user-specific spreading
code of length L.
L −1

∑c
1 (k ) (k )
Corresponding author is with the Department of Electronics and Communi-
cation Engineering, SSN College of Engineering, Chennai, India. E-mail: c (t ) =
l=0
l
p T c ( t − lT s ) (1)
prabagaranen@ssn.edu.in
2,4,5
Corresponding authors are with Ira A. Fulton School of Engineering,
Arizona State University. USA. E-mail: avswamin@asu.edu The pulse form of the chips is given by pTc (t). For the de-
3
Corresponding author is with Cognizant Techmology Solution,Chennai scription of the MC-DS/CDMA signal, the continuous
,India . E-mail:dhyaneshwaran.ks@cognizant.com time representation is chosen, since MC-DS-CDMA sys-
tems are of interest for the asynchronous uplink. Here,
Authorized licensed use limited to: VALLIAMMAI ENGINEERING COLLEGE. Downloaded on August 14,2010 at 15:40:25 UTC from IEEE Xplore. Restrictions apply.
2

OFDM might not necessarily be the best choice of multi- For unknown code channels, we do not know the signal
carrier modulation technique. power; thus, the path combiner output is given [3] by
The duration of a chip within a sub-stream is
L
1
N T
∑z
ˆ
T c
=T s
= c d
.
(2) uˆi ( v ) (m) = L
i, l
(v)
( m)αˆle − jθl (7)
L
(∑ αˆl 2 ) l =1

With multi-carrier direct sequence spread spectrum, each l =1

data symbol is spread over L multi-carrier symbols, each (for K + 1 ≤ i ≥ K + U )

of duration Ts .The complex-valued sequence obtained
after spreading is given by
4 PERFORMANCE ANALYSIS
N c −1
In the transmitter, QPSK and QAM are chosen as mod-
∑ d
(k ) (k ) (k ) j 2 π fn t
x (t ) =
n=0
n c (t )e , 0 ≤ t < L T s (3) ulation schemes. Spreading and scrambling techniques
are used for data spreading. Gold code is used for
spreading the user data which helps in user separation.
The received signal for user 1 is given by Turbo coding is employed as channel coding technique.
OVSF is employed as channelization code for robust
K
(4) transmission of data over multipath channel and the
y (t ) + ∑
T T
m
(t ) = P d (t )h
1 1 m
k =2
Pd k k
(t )h m(t )c1 c k + c1 n(t )
spreading factor of OVSF can be varied. The channel
I II III
TABLE 1
SIMULATION PARAMETERS
Where the first term in the above expression is the de-
Parameter Value
sired user signal, second term is the MAI and third term
is the noise Spreading codes (OVSF)

Scrambling code Gold code

3 PROPOSED MPIM SCHEME STRUCTURE Block size 256 (Variable)

Fig 2 elucidates the MPIM scheme. At the v stage, the th
Signal to noise ratio (Max) 10

Signal to noise ratio (Min) 2

Modulation QPSK and 16

QAM
Encoder rate ½ and ¾

Channel AWGN and

Rayleigh
model employed is AWGN and Rayleigh which provides
the upper bound and lower bound of system perfor-
mance.
-1
10

-2
10
Fig.2. MPIM Scheme
-3
matched filter output at the mth symbol associated 10
with the lth path for the ith code channel is represented by
BER

[3] -4
10
( m+1)Td +τl Iteration 1
zi , l ( v ) ( m ) = 1 ∫ rl ( v ) (t )ci N (t −τl )dt -5
10
Iteration 2
mTd +τl (5)
Td Iteration 3
Iteration 4
where Td(= N chips) is the symbol duration. -6
10
The despread signal is then coherently RAKE-combined 1 1.5 2 2.5 3 3.5 4
by using channel estimates. Considering the demodula- SNR(dB)
tion of QAM signals, the path combiner output for the
Fig.3. Performance of MPIM Scheme.
known code channels is expressed [3] as
L To enable statistically valid simulation results in reasona-
1
dˆi ( v ) ( m ) = ∑
ˆ
L
z i , l ( v ) ( m )αˆ l e − j θ l ble simulation time, Monte Carlo methods are used dur-
P i ( ∑ αˆ l 2 ) l =1
ing the simulation. Certain assumptions were used
l =1
throughout to simplify the simulation.Table 1 shows the
(for 1< i < K) (6)
simulation parameters considered for our simulation.

Authorized licensed use limited to: VALLIAMMAI ENGINEERING COLLEGE. Downloaded on August 14,2010 at 15:40:25 UTC from IEEE Xplore. Restrictions apply.
 3

The Fig 3 elucidates the performance improvement of the
MC DS/CDMA system for a set of data transmitted over
Rayleigh channel for various number of iterations. The
basic concept involved here is that, as we increase the
number of iterations the level of confidence of the MPIM
scheme in accession to turbo decoder to take a hard deci-
sion in favor of a 0/1 that was transmitted increases. For
instance, at a specific value of Signal-to-Noise Ratio, say
4 dB, the BER at the end of the first iteration approx-
-3.6 -5.5
imately dwindles from 10 to 10 at the end of the
fourth iteration. Fig 4 shows the image considered for
transmission in the simulation over AWGN channel.
The image shown in Fig 4 was transmitted over the
channel and subjected to the same decoding mechanism
as that of the data. The reconstructed image at the receiv-
er end is shown in the Fig 5 for various iterations. It has
been discerned that if the simulation trial is carried out
for a predetermined number of times, improvement in
the image received at the receiver swells. As it can be
seen from the Table 2, for seventh iteration the Symbol
Error Rate is very low compared to 1, 3 and 5 iterations.
5 CONCLUSION
In this paper, the performance of MC DS/CDMA has
been evaluated by incorporating the physical layer speci-
fications. The primary focus was to assay the perform
ance of MC DS/CDMA system with MPIM scheme over
AWGN and Rayleigh channel, to mitigate MAI. As antic-
ipated the system is interference limited. Several simula-
tions were carried out for the estimation of the perfor-
mance of MC DS/CDMA with spreading and scrambling.
Investigations were carried out on the Bit Error Rate and
SER of the downlink with the proposed MPIM scheme
for data and image respectively. It is discerned that the
new signal modeling along with Multipath Interference
Mitigation (MPIM) scheme can exterminate the unknown
intracell interferences .The investigation of simulation
results, divulge that the proposed scheme ameliorate the
pedestrian mobility needs to greater extent.

Fig.4. Transmitted Image

REFERENCES
[1] Kondo.S and L. B. Milstein, “Performance of Multi-carrier DS
CDMA Systems”, IEEE Trans. Comm., vol.44,no.2, pp. 238-246, Feb,
1996.I.
Iterations=1 and SNR=4dB Iterations =3 and SNR=4dB [2] D.Divsalar, M.K.Simon and D.Raphaeli “Improved paral-
lel interference cancellation for CDMA”, IEEE transc. Commun., Vol.46,
50 50 pp.258 – 268, Feb 1998.
100 100 [3] Arvind R ,Arun S, Boopathi N, Dhyaneshwaran K.S ,” A nov-
150 150 el Multipath interference mitigation technique for high-speed packet trans-
200 200
mission in WCDMA downlink’, Dissertation, SSN college of engineering.,
2007.
250 250
50 100 150 200 250 50 100 150 200 250 [4 ] Varun V Krishnan, Vasanthan V, Venkatesh R , “Space time
Iterations=5 and SNR=4dB Iterations =7 and SNR=4dB
processing to improve capacity of WCDMA system” Dissertation, SSN
college of engineering, 2006.
50 50 [5] “A Regulatory framework for UMTS.”, Report #1 from the UMTS
100 100 Forum, June 1997.
[6] Pieter van Rooyen, Michiel Lotter, Danie van Wyk , “Space-time
150 150
processing for CDMA mobile communications,” Kluwer academic publish
200 200
ers group, pp. 2-5, 2000.
250 250
50 100 150 200 250 50 100 150 200 250
[7] John G. Proakis, “Digital communications”, 4th ed. McGraw hill
series, 2001, pp. 518-521.

Fig.5. Reconstructed images

TABLE 2
SER FOR DIFFERENT ITERA TIONS WITH
SNR=4 DECIBEL
No. of Iterations SER

1 0.7434

3 0.4279
5 0.11
7 0.0084

Authorized licensed use limited to: VALLIAMMAI ENGINEERING COLLEGE. Downloaded on August 14,2010 at 15:40:25 UTC from IEEE Xplore. Restrictions apply.