Int. J. Communications, Network and System Sciences, 2010, 3, 540-547
doi:10.4236/ijcns.2010.36072 Published Online June 2010 (http://www.SciRP.org/journal/ijcns/).
Copyright © 2010 SciRes. IJCNS
Uplink Performance Evaluation of CDMA Communication
System with RAKE Receiver and Multiple Access
Interference Cancellation
Ayodeji J. Bamisaye, Michael O. Kolawole
Department of Electrical and Electronics Engineering the Federal University of Technology, Akure, Nigeria
E-mail: ayobamisaye@yahoo.com and kolawolm@yahoo.com
Received January 26, 2010; revised March 2, 2010; accepted April 10, 2010
Abstract
In CDMA communication systems, all the subscribers share the common channel. The limitation factor on
the system’s capacity is not the bandwidth, but multiuser interference and the near far problem. This paper
models CDMA system from the perspective of mobile radio channels corrupted by additive white noise gen-
erated by multipath and multiple access interferences. The system’s receiver is assisted using different com-
bining diversity techniques. Performance analysis of the system with these detect ion techniques is presented.
The paper demonstrates that combining diversity techniques in the system’s receivers markedly improve the
performance of CDMA systems.
Keywords: CDMA, Multipath Diversity, Multiple Access Interference Cancellation, Rake Receiver, Parallel
Interference Cancellation
1. Introduction
The mobile radio channel plays fundamental limitations
on the performance of wireless communication systems.
The transmission path between the transmitter and the
receiver can vary from simple line-of-sight to one that is
severely obstructed by buildings, mountains, and foliage.
Unlike wired channels th at are stationary and predictable,
radio channels are extremely random and do not offer
easy analysis. Even the speed of motion impacts how
rapidly the signal level fades as mobile terminal moves
in space. Modelling the radio channel has historically
been one of the most difficult parts of mobile radio sys-
tem design, and is typically done in a statistical fashion,
based on measurements made specifically for intended
communication system or spectrum allocation. Code
Division Multiple Access (CDMA) is a multiple access
technology that utilizes direct-sequence spread spectrum
(DS-SS) techniques. With this technology comes a para-
digm shift, including the use orthogonal or nearly or-
thogonal codes (so-called spreading sequences) to
modulate the transmitted bits. Contrary to the conven-
tional frequency division multiple access (FDMA) and
time division multiple access (TDMA) systems where
noise rejection d eals, primarily with out-of-band noise, a
CDMA system concerns mostly with inband noise. This
noise may come from self-jamming (or self-noise), back-
ground noise, man-made noise, inter-modulation, or
noise generated in the receiver [1,2]. If one can reduce
the unwanted in-band noise, such reduction translates
directly into improved performance. Undesired noise
comes from many different sources. This noise may
come from natural or human sources. Naturally occur-
ring noise includes atmospheric disturbances, back-
ground noise, and thermal noise generated in the receiver
itself. Through careful engineering, the effects of many
unwanted signals can be reduced [3]. DS-CDMA is a
multiple access system, where multiple users share a
limited resource, the frequency. Conventional asynchro-
nous DS-CDMA systems allow each user to transmit and
receive independently. Each receiver performs a simple
correlation between the received baseband signal and the
corresponding user’s spreading sequence. In a low-noise
channel with orthogonal spreading sequence,this app-
roach would be optimal. Due to the synchronicity of us-
ers and the need to support numerous users, such ortho-
gonality is impossible, even on a hypothetical AWGN
(additive white Gaussian noise) channel [4]. Thus, sys-
tem performance rendered multiple-access interference
(MAI) limited, and channel utilization is correspo ndingly
low.
This paper m odel s C DM A sy st em fr om t he pers pect ive
A . J. BAMISAYE ET AL.541
of mobile radio channels corrupted by additive white
noise generated by multipath and multiple access inter-
ferences. The system’s receiver is assisted using different
combining diversity techniques. Performance analysis of
the model with different detection techniques will be
presented.
1.1. Mobile Radio Channel Model
Mobile radio channel is one of the most important ele-
ments in the mobile communication systems. When the
signals are transmitted through Mobile radio channel, it
is affected by shadow or large scale fading. Mobile
communication is affected by multipath fading in addi-
tion to shadow fading. Multipath fading is caused by
atmospheric, scattering, and reflection from building and
other object. Multipath channel can be classified as dis-
crete (consisting of resolvable multipath components)
and diffuse (consisting of irresolvable multipath compo-
nents) [5,6]. Consequently, the multipath fading could
affect the transmitted sig nals in two ways: due to disper-
sion (also called time spreading or frequency selectivity),
and due to time variant behaviour of the channel (due to
motion of the receiver or changing environment such as
of foliage or movement of reflectors and scatter). This
means that the impulse response
,ht
of mobile radio
channel is time variant [6,7], and if
,
ht
has a zero
mean, then the en velope ht,

has a Rayleigh distribu-
tion with its probability density function described by

2
2
exp 2
rr
pr 2



(1)
where r is time dependent received signal (i.e.,
()
() ||
j
t
rtr e
, is its arbitrary phase) whose line of
sight (LOS), specular, and diffuse components are as-
sumed bounded in narrowband form, and 2
is the
total power in the multipath signal. In this case the total
power is considered to have zero-mean, amplitude fading.
Otherwise, if the impulse response has a non-zero mean,
then the envelope mobile radio channel would have a
Rician distribution with its probability density function
expressed as:


22 2
22
exp 2o
rs
r
prI 2
r

 




(2)
where 2
s
is the power of the line-of-sight component,
and o
I
denotes the zeroth order Bessel function of the
first kind argument (.). The Rayleigh distribution is a
special case of Rician distribution when s = 0; i.e., wher e
the LOS component is negligible.
In most general case, the channel can be modelled as
channel at time (t) to an impulse at time (t-τ). If x(t)
represents the transmitted (or emitted) signal through
noiseless mobile radio channels, the received signal, r(t)
can be expressed as

 
,rthtxtd



(3)
If the channel information is considered
ch , its distributed
annel impulse response may be expressed as
1N
 

0
,
j
j
j
ht t


(4)
In view of (4) in (3), we write modified channel re-
ceived signal:


1
0
N
j
j
j
rtxt j


(5)
β(t) is complex amplitude,
j is the path delay, and N
is
1.2. Rake Receiver
onventional matched filters are single pa th detectors. In
j
the number of multipath components. In the wideband
channel where the delay-line model had a large number
of taps, not all the multipath components are likely to
fade simultaneously. This may be used as a multipath
diversity to improve the received signal SINR (carrier to
interference plus noise ratio). Rake receiver, for instance,
can be used to mitigate the effect of fading if the trans-
mitted signal bandwidth is larger than the coherence
bandwidth; a practical example is in the wideband-
CDMA (WCDMA).
C
practice, when the transmitte d signal passes through mo-
bile radio channel, duplicates of the transmitted signal
are generated by reflection, refraction, and diffraction,
and the signal power is distributed in multipath. In
CDMA system, the transmitted signal bandwidth is much
larger than the coherent bandwidth of the channel, in
which case the channel is frequency selective [4,8,9]. For
the frequency-selective channels, the received signals are
multiple copies of the transmitted signals with different
channel delays and fading, combining the multipath
components as multipath diversity. Thus, if one of the
multipath components is attenuated by fading, some oth-
ers may not be and the receiver could use unfaded com-
ponents to make the decision. The idea behind the rake
reception technique is that the signals propagating
through different multipath are received in individual
fingers of the rake receiver and the outputs from these
fingers are then coheren tly co mbined to prov id e the inpu t
signal for the symbol decision. The received signal is
chip matched and sampled at the chip rate. Figure 1,
illustrates the structure of a typical rake receiver in which
*()
k
s
i and *,()
km
ci
(k = 1, 2, ..., K; m = 1, 2, ..., M) rep-
t, resply, the complex conjugate of chip- resenective
channel matched sampled signature sequence of the user
linear time-variant system [2] giving the response of the
Copyright © 2009 SciRes. IJCNS
A. J. BAMISAYE ET AL.
542
f interest and the complex conjugate estimate of the
n on the transmitted information bit is
ba
mate of the
n on the transmitted information bit is
ba
o
impulse response.
Then, the decisio
impulse response.
Then, the decisio
sed on the sum of the individual correlator’s outputs.
In our study, the dimension of each correlator equals the
system processing gain, and then the output of each fin-
ger

sed on the sum of the individual correlator’s outputs.
In our study, the dimension of each correlator equals the
system processing gain, and then the output of each fin-
ger

y
i is given after channel is matched and corre-
latin

y
g:
for
 
**
,,
111
PG MK
kmk jk
jmk
icisir i

 1
m
w
(where ) (6)
where PG gain,
pled sig
.2.1. Multistage Receivers r detector is the output of
the conventional matched filters receiver for single user k
In the case of multiple K active users, the received sig-
nal in the receiver is
So, the modified output of the co
filter receiver yk for the kth user is
(9)
Expression (9) consists of three terms. The first term is
the desired information which gives the signal of the
in
1, 2,,mM
is the processing *,()
km
ciis the complex
conjugate estimate of the channepulse response,
)(
*,is jk is the complex conjugate of chip-matched sam-
nature sequence of the users of interest.
l's im
1
The input signal of a multiuse
the matched filter bank, or rake receiver. Almost all
modern multiuser detection techniques deal with the out-
put of the matched filters bank and the cross-correlation
information of all users in the system. Expressions (1)
through (6) assumed noiseless cases, which is not valid
in real life situation. If the time-dependent interference
components,

nt , are assumed additive, the output of
can be expressed as:
 
** **
1
PG PG
kkkkkk
i
yicisiricisint


(7)
1i

() ()
K
k
rtr tnt
(8)
1k
nventional matched
  
**
1
PG
kkkk
i
yricisi
  
** **
11 1
KPG PG
jkk kk
ji i
jk
ri
cisi cisint



formation bit (which is exactly what is sought). The
second term is the result of multiple access interference
(MAI), and the last term is due to noise. The second term
typically dominates the noise so that one would like to
remove its influence. Its influence is felt through the
cross-correlation between the chip sequences and the
device-powers of users. If o ne knew the cross-correlations
and the powers, then one would attempt to cancel the
r(i)
)(
*is
k
)(
*is
k
)(
*is
k
c)(
*1, i
k )(
*2, ick)(
*,ic Mk
c
T c
T c
T
Correlator T
o(.)
w1
Finger1
Correlator T
o(.)
w2
Finger 2
Correlator T
o(.)
M
w
Finger M
y(i)
Figure 1. Rake Receiver with M fingers.
y1 y2 yM
Copyright © 2010 SciRes. IJCNS
A . J. BAMISAYE ET AL.543
in
front end, a SIC receiver (ordecorrelator) can be used
[1
on and Results
erformed for different
nce cancellation);
ting Detector and
M
is clearly stated. It is assumed that the system
us
ta is obtained by passing the soft
de
Simulations were performed for 10 active users sending
to 100total of 10000 symbols. The
in cance-
lli
Similar parameters used in 1.3.1 are used in this analysis
In add are updated ran-
.
Th
ectively, while other
pa
performance is the
sa
effect of one user upon another. This is, in fact, the
intuitive motivation for interference cancellation scheme.
The difference between multistage receivers and suc-
cesssive interference canceller (SIC) receivers is that
stead of using previous bit decisions to cancel interfer-
eence from desired user’s signal as in SIC, tentative de-
cisions on each user are used to improve signal quality
[10]. Receiver structure is called multistage, since when
decisions are made, they can be used to either make a
final decision on data or to enhance the signal through
cancellation. Reference signals are based on initial bit
estimates, which are then subtracted from received signal
to produce a cleaned spread signal for next stage. Since
all signals are detected at each stage simultaneously,
multistage receivers are also called parallel interference
cancellers (PIC). A two-user multistage receiver is
shown in Figure 2. The p-stage receiver outputs are

1,pp
k
yt
, where 1, 2,p.
Instead of a conventional (matched filter) receiver
a
1]. Performance of PIC is best when received signal
powers are equal. Capacity of the system is limited by
hardware.
1.3. Simulati
Performance evaluations were p
scenarios using:
1) Conventional matched filters receiver with PIC
(parallel interfere
2) Rake receiver with PIC;
3) Rake receiver with Decorrela
MSE; and
4) Conventional matched filters receiver and rake re-
ceiver.
For each analysis, the number of users’ symbols tran-
smitted
es unencoded BPSK signal, as well as no pulse-
shaping filter. Each transmitted MS within each loop
sends a block of data bits of known length. The input of
MUD (multi user detection) is the soft decision of either
rake receiver or conventional matched filters. In the
simulation, the sub-optimum linear MUD decorrelating
detector, linear minimum mean squared error (LMMSE)
detector, and nonlinear sub-optimum PIC were built for
the multishot model.
In relation to the decision operation, the hard decision
output of received da
cision output through the design circuit, which repre-
sents any sign function for BPSK. Also, we make the
number of channel paths P equal the number of fingers M
to make the rake receiver simpler in structure. In a less
equivalent case, that is, when M < P or M > P, the system
performance degrades. Therefore, the matched case pro-
vides the optimally achievable performance reference.
1.3.1. Performance of Conventional Matched Filters
Receiver with PIC
10 symbols within each loop. The maximum loop equals
. The users send a
data are then spread by Walsh code and scrambled by
Gold code with processing gain of 32. The channel pa-
rameters are updated randomly in each loop.
Simulation results are shown in Figure 3, which de-
monstrates progressive improvement in the system per-
formance with diversity combining techniques
ng multiple access interferences.
1.3.2. Performance of Conventional Matched Filters
Receiver and Rake Receiver
except the processing gain, PG, which is increased to 64.
ition, the channel parameters
domly in each loop and the number of paths for each user
changes randomly over (2-6). Two tests were carried out
in this subsection: (i) to investigate the effect of rake
receiver in the system for a single user and (ii) to inves-
tigate the effect of rake receiver and conventional
(matched filter) receiver with a range of channel paths.
Figure 4 shows the performance with rake receiver
and without rake receiver for a single user. In the re-
ceiver, ‘equal gain combining’ method is used in rake
e analysis demonstrates that lowering the SNR does
not improve the system when SNR equals 20 dB, the
BER is 10–2.9 and 10–1.9 respectively for conventional
matched filter receiver and rake receiver. In contrast,
improvement in system performance can be achieved by
increasing SNR in both sche m e s.
In the case of test (ii), an AGWN is assumed in the
channel paths. The number of users and the processing
gain are decreased to 6 and 8 resp
rameters are the same as previous simulations. The
channel parameters are updated randomly in each loop,
and the simulation computes for the different values of
channel path, as shown in Figure 5.
Figure 5 shows the performance of the rake and con-
ventional receivers for a number of multipath scenarios.
In the case of one path, the system
me for matched filter (MF) and rake receiver. But in
the case of multipaths; for example, (4-path) the system
performance with conventional matched filter degrades,
whereas the system performance with rake receiver im-
proves. Also, for the case of (7-path) the performance of
MF becomes very poor, but improves markedly with
rake receiver. As a result, the system performance can be
deduced as increasing when diversity scheme is em-
ployed. Under Gaussian noise as interferer, the system
performance improves with in creasing channel path with
the rake receiver, which might not be optimal in real life
situation.
Copyright © 2009 SciRes. IJCNS
A. J. BAMISAYE ET AL.
544
Delay +Conventional
receiver
Decision device
Sampled signal
generator
Sampled signal
generator
Decision device
Delay +Conventional
receiver
+
-
yk1
1t
+ -
yk1
2t
yk
1t
1()yt 1()
k
yt
1k
yk
2t
2()
k
yt
21()
k
yt
Figure 2. A two-user multistage receiver.
-20 -10 010 20 30 40 50 60
10-3
10-2
10-1
100
MF DD
MF MMSE
P IC - S tage1
P IC - S tage2
P IC - S tage3
Bit Error Rate
Figure 3. CDMA system performance evaluation using conventional matched filter (MF), with different diversity techniques:
decorrelating detector (DD), m mean squared error (MMSE) , and th ree-stage parallel internce cancellation (PIC). minimu fere
SNR (dB)
Figure 4. System performance of Rake receiver and conventional matched filters’ receiver.
Copyright © 2010 SciRes. IJCNS
A . J. BAMISAYE ET AL.
Copyright © 2009 SciRes. IJCNS
545
-20 -10 010 20 30 40
10
0
10
-2
10
-1
S y st em performance
Bit Error Rate
1 pat h : conve nt i onal m at ched fil t er
1 path :RA K E rec ei ver
4 path :c onvent i onal matc hed fil ter
4 path :RA K E rec ei ver
7 pat h : conve nt i onal m at ched fil t er
7 path :RA K E rec ei ver
SNR (dB )
Figure 5. System performance of Rake receiver and conventional matched filters’ receiver with different number of channel
paths.
1.3.3. Performance of RAKE Receiver with
Decorrelating Detec tor and MMSE
Similar parameters used in 1.3.1 are used in this analysis.
In addition, the channel parameters are updated ran-
domly in each loop, and the number of paths for each user
changes randomly between 2 and 5. Figure 6 clearly
shows the importance of multiuser detection (MUD). Teh
figure shows that for one path, marginal improve- ment in
system performance is gained whether conventional
matched filters receivers are used or in combination with
decorrelating detector (DD) or with minimum mean
squared error (MMSE) schemes. Note that the correlation
matrix only contains the information of multiple access
interferences in the first path and treated the other paths
as AWGN. On the otherd, system performance im-
at of [11].
1.3.4.ith Parallel
performance improved by using rake receiver instead of
conventional matched filters.
1.3.5. Perf or mance with Var iable Processing Gain
In this simulation, 8 active users send 10 symbols re-
spectively within each loop. The maximum loop is equal
to 100, totalling 8000 symbols sent by the users. The
data are then spread by Walsh code and scrambled by
Gold code with performance gains of 16 and 64. The
channel parameters are updated randomly in each loop,
and the number of channel path change randomly be-
tween 2 and 4. The received signal is processed with rake
and conventional matched filters respectively. Decorre-
lating detector is employed to cancel the multiple access
interference. Figure 8 indicates the processing gain in-
strate that large processing gains translate to high per-
CDMA system from the per-
pective of mobile radio channels corrupted by additive
interfver was assisted with
han
proves markedly using rake receiver and MUD when
taking the multipath into account; a finding consistent
ith th
fluence to the system performance. The results demon-
w
Performance of RAKE Receiver w
Interference Cancellation
For this simulation, 10 active users send 10 symbols re-
spectively within each loop. The maximum loop is equal
to 150, with a total of 15000 symbols sent. The data are
then spread by Walsh code and scrambled by Gold code
with performance gains, PGs, of 16 and 64. The chann el
parameters are updated randomly in each loop, and the
number of channel path change randomly between 2 and
4. The received signal is processed with rake and con-
ventional matched filter respectively. Three-stage paral-
lel interference cancellation is employed to cancel the
multiple access interference. As shown in Figure 7, the
formance gain of the system.
1.4. Conclusions
This paper has modelled a
s
white noise generated by multipath and multiple access
erences. The system’s recei
different combining diversity schemes.The simulation is
focused on the most important factors that will influence
the performance of the CDMA systems using multi user
detection method and interference cancellations scheme.
Performance analysis of the system using the different
detection techniques was presented. The paper estab-
Lished that diversity-combining techniques markedly im-
A. J. BAMISAYE ET AL.
546
-20 -1001020 304050 60
10
-4
10
-3
10
-2
10
-1
10
0
Sys t em pe
NR
e
rformance
conventional receiver
RAKE receiver
conventio nal receiver with DD
RAKE receiver with DD
conventio nal receiver with MMS E
AKE receiver with MMSE
R
Ra t
S
B i t Error
imum mean squared error (MMSE).
(dB)
Figure 6. System performan ce of Rak e receiver with decorrela
ting detector (DD) and min
-20 -10 010 20 3040
10
-3
10
-2
10
-1
10
0
E rr or ra te vs SNR
B i t E rror Rate
conventional receiver
RAKE receiver
MF:PIC - Stage1
MF:PIC - Stage2
MF:PIC - Stage3
RAKE:PIC - Stage1
RAKE:PIC - Stage2
RAKE:PIC - Stage3
SNR ( dB)
Figure 7. The system performance of RAKE receiver with three-stage parallel interference cancellation (PIC).
-20 -10 01020 30 40 50 60
10
-4
10
-3
10
-2
10
-1
10
0
S y st em pe rformance
S NR (in dB)
B i t Error Rate
conventional receiver(PG=16)
RAKE receiver(PG=16)
conventional receiver with DD(PG=16)
RAKE receiver with (PG= 16)
conventional receiver(PG=64)
RAKE receiver(PG=64)
conventional receiver with DD(PG=64)
RAKE receiver with (PG= 64)
Figure 8. System performance of RAKE receiver with differen
SNR (dB)t.
Copyright © 2010 SciRes. IJCNS
A. J. BAMISAYE ET AL.547
prove t
portant factor in reducing the multiple access interfere-
ence, is the processing gain which the model employed.
2. References
[1] D. Torrieri, “Principle of Spread-Spectrum Communica-
tion Systems,” Springer Science + Business Media, Bos-
ton, 2005.
[2] R. Prasad, “Universal Wireless Personal Communica-
tions,” Artech House, 2000.
[3] S. C. Yang, “CDMA RF System Engineering,” British
Library Cataloguing in Publication Data, II. Series,
TK5103.2.Y36, Irvine, California, 1998.
[4] A. J. Bamisaye, “Effect of Multiuser Interference on
Subscriber Location in CDMA Network,” M Eng Thesis,
Department of Electrical and Electronics Engineering, the
Federal University of Technology, Akure, 2009.
] A. J. Bamisaye and M.
] T. S. Rappaport, “Wireless Communications Principles &
02.
[7] M. Frikel, B. Targui, F. Hamon and M. M’Saad, “Adap-
tive Equalization Using Controlled Equal Gain Co mbining
for Uplink/Downlink MC-CDMA Systems,” International
Journal of Signal Processing, Vol. 4, No. 3, 2008, pp.
230-237.
[8] A. Kansal, S. N. Batalama and D. A. Pados, “Adaptive
Maximum SINR RAKE Filtering for DS-CDMA Multi-
Path Fading Channels,” IEEE Journal on Selected Area
in Communication, Vol. 16, No. 9, 1998, pp. 1765-1773.
[9] T. Kim, K. Ko, Y. Kim and D. Hong, “Performance
Evaluation of Uplink MC-CDMA Systems with Residual
Frequency Offset,” IEICE Transactions on Communica-
tions, Vol. E89-B, No. 4, 2006, pp. 1455-1458.
[10] K. Vardhe, D. Reynolds and M. C. Valenti, “The Per-
formance of Multi-User Cooperative Diversity in an
Asynchronous CDMA Uplink,” IEEE Transactions on
Wireless Communications, Vol. 7, No. 2, 2008, pp. 1930-
k
ty
he performance of CDMA systems. Also, an im- Practice,” IEEE Press, New York, Prentice-Hall, 20
[5 O. Kolawole, “Evaluation of 1940.
[11]
Downlink Performance of a Multiple-Cell, Rake Receiver
Assisted CDMA Mobile System,” Wireless Sensor Net-
work Journals, Vol. 2, No. 1, 2010, pp. 1-6.
[6
X. B. Yu, X. D. Zhang, D. Z. Xu and G. G. Bi, “Uplin
Performance of CWP-MC-CDMA with Space Diversi
Combining Technique over Rayleigh Fading Channel,”
Journal of Electronics, Vol. 23, No. 6, 2006, pp. 837-841.
Copyright © 2010 SciRes. IJCNS