Performance Of Multi Tone Code Division Multiple Access (mt-cdma) In An Awgn Channel And In Presence Of Narrowband Jamming

Performance of Multi tone Code Division Multiple Access (MT-CDMA) in an AWGN Channel and in Presence of Narrowband Jamming Khalifa Nasser K Jileta, Mahamod Ismail, Alina Marie Hasbi Department of Electrical, Electronics and System Engineering, Faculty of Engineering Universiti Kebangsaan Malaysia 436000 Bangi- Selangor, Malaysia Tel: +6 03 8296322 Fax: +6 03 8296146 kjleta(yahoo.com, {mbi, alina)}eng.ukm.my

Abstract- The main limits of the conventional single carrier modulation techniques are the restrictions imposed by the multipath channel, and the receiver complexity. On other hand the multicarrier techniques such as MultiCarrier (MC) and Multi-Tone (MT) can provide high data rates at reasonable receiver complexities. Jamming on the other hand is of interest in some communication applications, and in military anti-jam systems. In this work, we study the performance of Multi-Tone Code Division Multiple Access (MT-CDMA) in presence of narrowband jamming. We investigate Bit Error Rate (BER) performance as a function of some system parameters such as number of sub-carriers, processing gain, and as a function of channel conditions such as jamming type, number of jammers, jamming center frequency, jamming power, and jamming to signal power ratio. We compare the performance with the analytical performance in absence of any jamming. The simulation results show, that the performance can be enhanced by increasing the processing gain, number of jammers or by decreasing the number of sub-carriers, the jamming to power ratio or jamming power. Furthermore the effect of jamming center frequency on the performance is negligible, and good agreement with the analytical performance in absence of jamming is achieved.

Keyword&-OFDM, CDMA, MT-CDMA, Narrowband

jamming

I. INTRODUCTION

Multimedia applications of mobile communications need modulation, and multiple access techniques that, can deliver very high data rates, which cannot be offered. by the traditional single carrier modulation techniques. MTCDMA is a combined technique between Orthogonal Frequency Division Multiplexing (OFDM) and Code division Multiple Access (CDMA) and it is an attempt to provide a communication system that inherits the advantages of both CDMA and OFDM, and provides the new probable technique for the fourth generation of mobile communication systems. MT-CDMA inherits the powerful features of CDMA, which allow number of users to access the channel simultaneously. This is achieved by, modulating and spreading the signals with pre-assigned codes sequences. Since CDMA provides better spectral efficiency and easier base station placement compared to second generation systems and beside its high perfonnance in presence ofjamming interference. Another

1-4244-0000-7/05/$20.00 C2005 IEEE.

very important advantage of MT-CDMA gained from OFDM is its less sensitivity to the inter symbol interference (ISI) caused by radio channel impairments. This advantage is achieved by using more than one carrier to carry the symbol and then a lower data rate for the same user data comparing to single carrier modulation is achieved. By well-chosen system parameters, we can reduce the effect of unfriendly channels such as frequency-selective channels. Another very important advantage using MT-CDMA is its spectrum efficiency due to allowing the sub-carriers to largely overlap with each other. Actually High Spectral Efficiency is the main advantage of MT-CDMA over single and the multicarrier techniques. At the same time, to facilitate inter-carrier interference free demodulation of the sub-carriers, the subcarriers are made orthogonal to each other. The possible orthogonality in MT-CDMA is over the symbol time T, [1], [2]. If the orthogonality is not altered by the channel, the modulating signal of each sub-carrier can be recovered exactly by the receiver [3]. A comparison between MT and single carrier schemes was done in [4] and both schemes show equivalent performance in presence of multi user interference. Jamming noise is of interest in some fields and applications such as anti-jam systems used in military applications and in the spectral overlay of narrowband pulsed signals over Direct Sequence Spread Spectrum (DS-SS) transmission [4]. In this work we assume single user to simplify the work (no Multi User Interference) and we use conventional receiver which is recommended in case of single user detection. Spreading technique is DS-SS and the modulation technique is assumed Binary Phase Shift Keying (BPSK), since it is the most common modulation type, used in direct sequence systems. The type of data mapping to sub-carrier is considered as Serial to Parallel (S:P) transformation. The codes on all sub-carriers are the same, in order to obtain a high correlation among the sub-carriers modulation symbols. Perfect symbol synchronization and phase coherence are assumed. And the channel is assumed as Additive White Gaussian Noise (AWGN). In the next we describe MT-CDMA system. In section III we describe the simulation methodology. Section IV provides the simulation results, discussion and section V includes a summary and the conclusion.

979

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY MADRAS. Downloaded on May 28, 2009 at 21:20 from IEEE Xplore. Restrictions apply.

11. SYSTEM DESCRIPTION ANT MODEL Fig. 1, Fig. 2, show the MT-CDMA transmitter and MT-CDMA spectra respectively.

(t)

Cos(2*pi*fxT)

C (t)

Cos(2*pi*f2T)

C

S

{

j

, x

Ij

j QC(n)cos(2Tj,t)4 (3)

where: S(t) is the transmitted signal. M is the number of sub-carriers. is the integer part of x.

Lxj

d(t)

cos 2z'rfjT(t): is the RF carrier for the corresponding sub-carrier at the transmitter. At the receiver, the transmitted signal is added to n(t), the AWGN noise signal with zero mean and variance of NJ2 W/Hz. The received signal can be

S(t) *00

I I I I CM(t) Cos(2*pi*fMT)

r(t) = s(t) + n(t)

Fig. 1. MT-CDMA Transmitter

'- I I I I ! f2 f3 f3...fM

f1

Fig. 2. MT-CDMA spectra

The incoming bit stream denoted by d(t) is first serialto-parallel converted into M data streams (Mis number of sub-carriers). Then the data on each path is spread by using orthogonal codes denoted by Cj(t).Same or different codes can be used on each path. After spreading the spread data modulates the orthogonal sub-carriers. In MTCDMA, the sub-carriers are orthogonal over the symbol duration, and hence the sub-carriers frequencies are given by f

=

f

where

+P

fo

(4)

MT-CDMA receiver is shown in Fig. 3. The received signal is entered to M parallel paths, on each path we demodulate the received signal by multiplying it by cos 2fjR (t) and c(t), then integrating over the symbol period (Ts). The integrator removes (shown in Fig. 4) the double or high frequency term resulting from the multiplication by the sinusoidal carrier signal. The decision circuits are used then to collect the symbol samples to make a decision and get the data symbol. The decision static is denoted by 4. This is done on each subcarrier, hence Mof Fig. 4 is required [5]. Cos(2*pi*f,T)

is the RF and T is the

01,.....M-l. The modulated and symbol time, and p spread waveforms are scale and then summed. The Jlh sub-carrier data waveform (transmitter input) is[5]: =

dj(t)

=

2 n

djn

p

t-n

(l)

djn

E D = {1}and T is where the nib data symbol is the bit time. The modulation symbol Ij(n) are analogous to the data symbol and the modulation waveform for subcarrierj can be defined as[5].

Ij (t) =EIj (n)pTs (t- nTs)

Fig. 3. MT-CDMA Receiver

(2)

1.(t)

The Processing Gain (P) is defined as P = T/TC, where T, = 1/R1 the chip time, and R& is the chip rate. The modulated and spread waveforms are scaled and then summed. The final carrier waveform is denoted as S(t) [5]

Ts

_(t)

(t) Fig. 4 De-spread block of MT-CDMA receiver

C

980

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY MADRAS. Downloaded on May 28, 2009 at 21:20 from IEEE Xplore. Restrictions apply.

MT-CDMA analytical model is given in [5]. [5] Assumes AWGN and doesn't consider any jamming interference. The AWGN channel receiver decision statics Zj is given as [5]

Ts

Z= r(t)cos(27dtRt)c(t)dt

each sub-carrier and all the sub-carriers samples are then summed to perfonn the transmitted signal given by (3)F but for one user (1 =1). D. AWGNchannel Rand Matlab function is used to generate AWGN. The variance of AWGN is changed with the desired Eb/NO. The transmitted signal is then added to the AWGN to get the transmitted signal with AGNW noise. That signal is then added to the jamming noise to perform the input signal to the receiver.

(5)

0

The probability of error was derived in [5] for BPSK and it is given by

PB

=

QEaAjJ

(6)

Equation (6) is used to compute the BER analytically for MT-CDMA in absence of jamming interference and it is used in this work to compare the performance in presence and absence of jamming interference. 1Il. SIMULATION DESCRIPTION

We have employed Monte Carlo simulation technique to estimate the performance of the system in terms of probability of error. Monte Carlo simulation method utilizes sequence of random numbers to perform the simulation. To study the MT-CDMA system performance, a software model of the system is developed (transmitter, communication channel, and receiver). The simulation program started by setting the simulation parameters and then a random data is generated. The simulation model can be summarized as follows: A Serial to parallel (S: P) Converting the serial data to parallel is a simple process. Dividing the input data with rate Rb bps to M parallel streams each of which is Rb/M bps where M is the number of sub-carriers does the whole process. B. Spreading code generator The code that, used to spread the data is also randomly generated The spreading process is done based on DS-SS technique, where each data bit is repeated first P times, then randomized by multiplying it by a random number (pseudorandom code) generated above. By repeating each bit P times, we achieve the first condition for the DSCDMA signal; we get chips with chip time equal to Tb/P or R, equal to P*Rb (spread the data over wider bandwidth), and then by multiplying the chips with random number we achieve the second of DS-CDMA signal which seems random to the jammers. C. Sinusoidal Carriers Generators Sinusoidal sub-carriers generators are used to generate sinusoidal vectors for modulation. The sub-carriers of MT-CDMA as mentioned before are separated by l/(T,) in frequency domain, and since we have one sample per chip and each spread bit is sampled P times, then in simulation the sinusoidal vectors are separated by I/P. The data is multiplied by the spreading code and sinusoidal signal for

E. Jamming genera tor Jamming noise can be considered as sinusoidal signals. Both of single tone and multiple tone jamming noises can be generated by the same method used to generate the sinusoidal carriers.

F. Detection and error estima tion The receiver is simulated based on the system shown in Fig. 3. We use conventional receiver for the detection process, which consists of correlator, integrator and comparator. The received signal enters to the receiver is first serial to parallel converted into M parallel paths. At each path it is demodulated and despread by the correlator by multiplying it with the sinusoidal signal and spreading code. After that it is integrated or accumulated over the symbol time, where the double frequency term resulting from the multiplication by the carrier signal is removed. Then, decisions are made on the bits by the decision circuit shown in Fig. 4 after collecting the symbol samples and translating them to data symbol based on threshold value (greater than or equal to 0.5 is considered 1, and less than 0.5 is 0). To calculate the errors exist in the received data, we compare these received bit estimates with the transmitted bits. In other words, the bits of each transmitted subcarrier are compared with the corresponding received subcarrier bit. The error counter is also increased if the transmitted and received bits are not the same. The bit error of the all system (aggregate bit error rate) is then the average of these bits error of all sub-carrier. The output of this simulation program is a plot of the obtained bit error of the system versus SNR (ENo). IV. SIMULATION RESULTS

A Performance in Presence ofSingle Janming The most important parameter that effect design of MTCDMA is the number of carriers, since in MT-CDMA the bandwidth is divided among large number of sub-carriers, and the sub-carriers are largely overlapped. Fig. 5 presents the effect of number of carriers on the system

performance.

981

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY MADRAS. Downloaded on May 28, 2009 at 21:20 from IEEE Xplore. Restrictions apply.

1n0

0peformnance of MT-CDMA in presence of single jamming over AWGN channel . .::.::-:_-.: ::-. :.t-. -_:I _ _ ,_ F_ _ _-, :=::::

-=

_-

,

_ ___ _

___

--

==-- -,=^ts., s

=R ---

.

_

-------

--

lo-'I

.

_F .

b-<------f---

=,=s|t-=sr.

-T---------------

-

AWGN

*---------

~---~ _-_- _-_-__-_-_-,

10Uo-2

s.

. . . ._

10,~

--

--

-----

__ __. __-----------

lo'

---

__ __ _ , ~~-<

--~~-- ~_

----

-

t

-----

- -

*---------T -- ----,-------,--

11 0 1 2 3

4

r

S

:-

-

:

-'-* + * --- ---;- -------

---

----

----

5

SNRtdB)

-

___

6

,__

_~,___

8

7

9

10

Fig. 5. Effect of number of sub-carriers

Fig. 5 presents clearly that by increasing number of carriers, the BER will increase, due to the increase in the sub-carriers interfering, and hence the system performance will be degraded. The result is based on three simulations performed with different number of carriers (3 carriers, 5 carriers and 15 carriers) for various SNR (O dB to 10 dB). Indeed some factors may limit the number of carriers in the system, such as sub-carriers interfering, system complexity and processing time. With small number of sub-carrier (M13), the performance obtained was very close to the theoretical BER in absence of any jamming. We can also notice that the effect of single jamming noise is negligible due to the effect of corrleator and the integrator where the jamming is first spread over the entire system bandwidth by the correlator and then lowpass filtered by the integrator so, that only small portion of its power will be collected. The effective power of the jammer is reduced by a factor equals to the processing gain [7]. The effect of processing gain is showrn in Fig. 6. performance of MT-CDMA in presence of single jamming over AWGN channel ~~~--------4

.i.~ ~ ~

:::.1~~~~~~~~--

----

-----

----

10

.

P-32 -P=64

------ ------1 ---

i -i-El -------------------------Li

10

-----------I ----------------

-----------------------

10m

--------------

............I

-------------

--------------

.............

---------------------

--------------

--------------

......

.................... -----------

-el in-

--------------

I

0

1

2

3

4

5

SNR(dB)

IE

6

--------------

7

8

9

10

Fig. 6. The effect of processing gain in presence of single tone jamnming noise

Dperformance of MT-CDMA in presence of single jamming over AWGN channel -E

ti-iE-ti-ii

t-----t-iE

-i-i--- --- E--E----X

and over AWGN channel

vI

--,* -

J-S--,,*,}9dB.

oe ca Anel AWN..................... the simulation resu. , It i-s cle ~~~~~~~~~~~~..

tone jm i

investigated. F-ig. 8--sho

..........................

-----s~~--~

~

SNR(dB)~~~~~J S-d

t

push

er To be ce---at -e ta -t

th

B'.PEFRAC .... AWG ..... CHNE ....... .. . ..3... 3sE..AN iN............. ..... OVE -

1

2

3

4

5

6

7

8

9

10

SNR(dqfro Fi 8 tha, inraigtenubro_amn 7. The of. jamming to signal power ratio in presence of single effect Fig. tone jamming over AWGN channel

desre jamming csrs:aseand moSs=2sts:of-deisis$sgner'2}sdo so ffrt B. PERFORM4ANCE OVER AWGN CHAMFL AND) IN PRESENCE OF MULTIPLE TONE JAMM7ING In this part number effect side of so the thethe receive effectis jammers cOrrelatorr ratr th=ren tha:rer=jofammin investigsed. Fig. 8 shows the simulation results. It is clear from of jamming increasing redceFig.by 8thethat,inert. Tistheis number tru if thiamn signals enhances the performance. This is because as we increase the number of jamming noises as we go toward white wide band jamming case which is actually the desired jamming case and most of designers do so efforts to push the jammer to be at that case because the total jamming power will be divided among a number of jammers and then spread over a wide band by the correlator at the receiver side so then the jamming effect will be spread over all the sub-carriers and then will be reduced by the integrator. This is true if the jamming power is not sufficient enough to interrupt the transmission, as shown in Fig. 9. g

P =128 AWGN

...........

achieved by using higher processing gain but the cost of increasing the processing gain is less throughput, lower bit rate, and higher system complexity. The effect of jamming interference comes from its center frequency and its power with compared to the signal power. Maximum jamming effect is achieved if the tone is placed on one of the sub-carrier center frequencies [6]. Fig. 7 shows the effect of single tone jamming center frequency and jamming to signal power ratio. First the jammer was replaced at the center of the first sub-carrier, and then at the center frequency of the transmitted signal. As shown in Fig. 7, it is clear that, the jamming has no effect on the transmission even if it has centered frequency as one of the sub-carrier center frequencies (worst case). However, by increasing the jamming to signal power ratio from 3 dB to 9 dB, the jamming noise has clear effect on the performance of the system.

It is clear from the results in Fig. 6 that increasing the processing gain can enhance the system performance. Higher capacity in terms of number of users can be also

982

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY MADRAS. Downloaded on May 28, 2009 at 21:20 from IEEE Xplore. Restrictions apply.

x~r-

Replacing the single jamming at the center frequency of one of the sub-carriers frequencies has a negligible effect and almost the same as replacing it at the center of the spectrum of the transmitted signal however, the performance is degraded as jamming to signal power ratio increases. Increasing the number of jammers will enhance the performance, but increasing jamming power degrades the performance. Excellent agreement between performance in AWGN in presence of jamming and the theoretical plot of the performance in AWGN without any jamming.

peorrnance of MT-CDMA in presence of Mutiple tone jamming over AWGN channel Jammers= M -, _l --i 3 t__:- .E . - - I_ _ = iE _-HE*_-E ---. jammers=NM _ -- 9- jammers=2M _

----------

-------

------ T------ I

------- --

it 10

_ 5_ _t--

...............

-3--------------------÷

cc

it

w

-----

~

~

~

----

-

jammers=4M

-

AWGN

,,

~

-

_.___i______*______J__.

b

---

-

--------------=

------=: -------t

-

_ _ _ _ _* _ _ __ _ _ __ _ . __ .

30

it

-10.01 0

it

........

.

.

F__-----

......-

..

-----;- -- -- -;-- -- -- -,--- -- - - r

i --------- --

- - - - - -

-- - -

--

-- --

---

5

l 3

2

1

------------

5

4

6

-r---:

.

: -: -r;------

______ >Q ______

VI. REFERENCES

10

9

8

7

SNR(dB) Fig. 8. The effect of number ofjammers over AWGN channel

[11 D. W. Matolak, V. Deepak, F. A. Alder, Perfornance of

Fig. 9 represents the effect of increasing the jamming power on the system performance. The result is based on three simulations perfonned with jamming power (equal to the transmitted power, double of the transmitted power, triple of the transmitted power). From Fig. 9, we can easily notice that the BER of the system decreases as the jamming power increases even when the number of jammers is higher than the number of the sub-carriers so we can conclude that the effect of multiple jamming noise is controlled by its power and the increasing of the number of jamming signals can enhance the performance of the system unless the jamming power is not sufficient enough to interrupt the transmission. -peformance of MT-CDMA in presence of Mutipte tone jamming over AWGN channel

.,------,---------

.

_ _ -_. _ S_

__

10

ci

tu 30

___

s-

- - ----- ~e

r. ---

_______-------r------

-- --

J=2TSrans.power J-3

-.-:

----

~-- -

-

_ -- -- ~---------T----- . - -----------__

-----

--

-

:-

10-t 10.1 0

1

2

_

-----___=_____r__

-

r---

_

1997. [3] L.Vandendorpe, Multitone Spread Spectrum Multiple Access Communications System in a Multipath Rician Fading Channel. IEEE Transactions on Vehicular AR technology. Vol. 44. NO. 2. 1995. 14] D. W. Matolak, F. A. Alder, V. Deepak, Performance of Multitone and Multicarrier DS-SS in The Presence of Partial-Band Pulse Jamming/Interference. School of Electrical Engineering & Computer Science, Athens, OH 45701.2002. [5] 1. Sen, Bandwidth Efficient Reduced-complexity MT-DS-SS via Reduced subcarnier Frequency Spacing. M.Eng. Thesis. Carleton University. Ohio University. 2004. [6] R. L. Peterson, R. E. Ziemer, D. E. Borht, Introduction to Spread Spectrum Communication, Prentice-Hall, Upper Saddle River, New Jersey, 1995. [71 Tan. F. Wong. (2004, May 12). Spread Spectrum and Code Division Multiple Access. Wireless communication course. University of Florida. Available:

http://www.wirelesss.ece ufl-edu/twong/eel6503/

Trans.power

-

Multitone and Multicarrier DS-SS in the Presence of Imperfect Phase Synchronization. School of Electrical Engineering & Computer Science, Athens, OH 45701.2002. [21 S. Hara, R. Prasad, Overview of Multicarrier CDMA, IEEE Communication Magazine, vol. 35, no. 12, pp. 126-133, December

*-*

-

%

--* _-R

*

7----------:: C = =r ra------------ ----------:: ------:: : ;: : :: : : ~: ------- ------

3

4

5

SNR(dB)

6

7

a

9

10

Fig. 9. The effect of the power of multiple tones janming noise over AWGN channel V. CONCLUSION

In this work, the performance of MT-CDMA under various channel, various system conditions and in presence of narrowband jamming noise has been investigated. Increasing the number of sub-carriers will degrade the system performance. The sub-carriers interfering limit the number of sub-carriers. Increasing the processing gain enhances the system performance.

983

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY MADRAS. Downloaded on May 28, 2009 at 21:20 from IEEE Xplore. Restrictions apply.