ber performance improvement using fading techniques

7/29/2019 BER performance improvement using fading techniques

http://slidepdf.com/reader/full/ber-performance-improvement-using-fading-techniques 1/5

B

E

Abstract —

multipath fadproblems in p

channel and tthe receiver. T

due to differ

channel andalgorithms for

of Additive W

are explored

particular, Le

equalizers arcompensate fo

of equalizers i

we have combCombing Dive

combining teinterference a

approaches h

performance.

Key Words

The radio

multipath fad

interference iinterference f

diversity algocan be para

characterizedamplitudes sh

of terminals

large number (OFDM) sym

estimating an

the wireless

attenuation, aof OFDM is

function of c

amplitude anfrom the av

removes the

subsequent sy

OFDM is a

can accommo based wirele

integral part

MIMO-OFDdivides the av

but orthogo

R Per

viron

The wireless c

ing characterioper reception

e type of fadihe channel esti

nt types of i

in equipmendigital comm

hite Gaussion

and their p

ast square Er

used to pror Inter-Symbol

variable in m

ined Equal Garsity techniqu

hniques is ad Inter-symbo

ave resulted

OFDM, Equ

I. INTR

hannels in m

ing channels,

the receivedrom the signa

rithm can beeterized as

by a delayow fast tempo

hile the del

of orthogonal bols. A possib

d tracking the

channel cause

d phase shiftto mitigate th

annel estimati

phase shiftailable pilot

effect of the

mbol demodul

special case o

date high datas systems. S

of OFDM s

systems proailable spectr

al narrowba

orma

ent Us

Deepmal

mmunication

stics and the. Therefore th

g is most impmation has bec

nterference pr

ts. In thisnications syste

oise and Mult

erformance is

ror equalizer

vide the optierror. As the

ultipath fadin

in combining as, and find t

ble to fightl interference

in big impr

lizer, Diversity

DUCTION

obile radio sy

which are ca

signal. To rel, much kind

sed. The wire

a combinatio

and compleal variations

ys are almos

frequency divle approach i

delay [1].it i

s an arbitrar

in the receiveffect of ti

on is to form

caused by thinformation.

wireless ch

ation.

multicarrier t

rate requiremince channel

ystems. As

mises higher m into a num

d sub chan

o ri ht © 2

ce Imp

ng Co

TSingh Pari

is impaired by

efore createsknowledge o

rtant in desigome very vast

esent in wire

thesis, estimams in the pres

ipath environ

investigated.

and Zero for

um solutionBER perform

channel there

nd Maximal Rat Maximal R

with Co-Charoblem. The a

vement in

, QAM.

stems are usu

sing intersy

ove intersyof equalizers

less radio chaof paths,

amplitude.ue to the mob

t constant ov

sion multiplethat of expli

well known

time dispers

d signal. Thee dispersion.

an estimate o

e wireless caThe equaliza

nnel and all

ransmission a

ent of multimestimation is

combinatio

ata rates. OFer of overlap

nels, and h

Internation

012 IJECCE,

rovem

bined

chniqar

the

thethe

ningfield

lesss

tionnce

ent

In

cing

andnce

fore

atioatio

nnelove

ER

ally

bol

boland

nelach

Theility

er a

ingitly

that

ion,

useThe

the

neltion

ows

d it

ediaan

of

DMing

nce

conver

selecti

OFD

Transf

subcar comm

capabil

with r bandw

examp

hiperlaOFD

subjec

OFD

Modul

in theMultip

other i

FDM.

indepemultipl

Fig. 2.

PN

l Journal of Ele

ll ri ht reserv

nt in

Equali

esProf. R

ts a frequency

e channel [2].

ORTHOGO

M is a mult

rm/ Fast Fo

iers. OFDM inications sys

ity with high

gard to multdth is divide

e 2000-8000

n2 and data ismethod is so

to multipath t

M is a combi

ation means a

arrier phase,lexing means

dependent da

In OFDM

dent channeexed to create

Block diagra

equence and

tronics Commu

V

ed

ultipa

zer an

vi Mohan

selective cha

II. OFD

AL FREQU

ULTIPLEX

icarrier syste

rier Transfo

s becomingems due to i

andwidth effi

ipath fadingd into very

hz for digi

transmitted i popular for ne

ransmission.

Fig. 1.

ation of mod

mapping of th

requency or aa method of s

ta channels. O

the signal it

s, modulatedthe OFDM ca

of an OFD

guard bit ins

ication and Co

olume 3, Issue 2,

h Fadi

Diver

nel into a no

ENCY DIVIS

NG)

uses Discr

m and

idely applieds high rate tr

ciency and its

nd delay [4].any narrow

al TV and

parallel on tw broadband

ulation and m

e information

mplitude or charing a band

FDM is a spe

self is first

by data anrrier.

system using

rtion FFT is

161

puter Engineeri

ISSN 2249 –071

g

ity

frequency

ION

te Fourier

spectra for

in wirelessansmission

robustness

Available bands for

48 hz for

ese bands.ystems are

ltiplexing.

on changes

mbination.width with

cial case of

split into

then re-

FFT, pilot

ritten as

g

X



162

International Journal of Electronics Communication and Computer Engineering

Volume 3, Issue 2, ISSN 2249 –071X

Copyright © 2012 IJECCE, All right reserved

∑ / , 0 1

…. (1) W N be the complex-valued phase factor

/

Thus, X (k) becomes

∑

, 0 1

…. (2)

Similarly IFFT is written as,

∑

, 0 1

.… (3)

III. QUADRATURE AMPLITUDE MODULATION

(QAM)

QAM is more efficient in terms of bandwidth than either

FSK or QPSK, but it is also more susceptible to noise. Thedisadvantage of DSB signals which is occupy twice the

bandwidth required for the baseband can be overcome by

transmitting two DSB signals using carriers of the samefrequency but in phase quadrature [5].

In this figure, the boxes labeled – π/2 are phase shifters,which delay the phase of an input sinusoid by – π/2 rad. If

the two baseband signals to be transmitted are m1 (t) and m2

(t), the corresponding QAM signal φQAM (t), the sum of the

two DSB-modulated signals is

….. (4)

Both modulated signals occupy the same band. Yet two baseband signals can be separated at the receiver by

synchronous detection using two local carriers in phase

quadrature.

2

2 ……... (5)

2 2 ……... (6)

The last two terms are suppressed by the low-pass filter,

yielding the desired output . Similarly, the output of the

lower receiver branch can be shown to be . This scheme isknown as quadrature amplitude modulation (QAM). Thus, two

baseband signals, each of bandwidth B Hz, can be transmitted

simultaneously over a bandwidth 2B by using DSB transmission

and Quadrature multiplexing.

IV. EQUALIZER

Theoretically, an equalizer [6] should have a frequency

characteristic that is the inverse of that of the transmissionmedium. This will restore higher frequency components and

eliminate pulse dispersion. Unfortunately, this also increases the

received channel noise by boosting its high-frequencycomponents. For digital signals, however, complete equalization

is really not necessary, because a detector has to make relativelysimple decisions- such as whether the pulse is positive or

negative.

4.1 Zero Forcing equalizer (ZF)It is really not necessary to eliminate or minimize ISI with

neighboring pulses for all t. all that is needed is to eliminate or minimize interference with neighboring pulses at their respectivesampling instants only, because the decision is based only on

sample values. This can be accomplished by the transversal-

filter equalizer encountered earlier, which forces the equalizer output pulse to have zero values at the sampling instants. In

other words, the equalizer output pulses should satisfy the

Nyquists criterion or the controlled ISI criterion.

1

……… (7)

4.2 Least Mean Squared Error Equalizer (LMSE)Another approach to equalization, the least mean squared

error method, does not try to force the pulse samples to zero

at 2N points. Instead the mean of the squared errors over a

set of output samples is minimized [7].

The prediction error is given by

^

^

…………. (8)

…………. (9) W k is a weight vector

To compute the mean square error || at time instant k.

||

2

……….… (10)

Taking the expected value of || over k yields

||

2

…. (11)



163




W k, filter weights are not included in the time average since,

for convenience, it is assumed that they have converged to

the optimum value and are not varying with time.

V. DIVERSITY

Diversity techniques [8] are based on the notion thaterrors occur in reception when the channel attenuation is

large, i.e. when the channel is in a deep fade. If we can

supply to the receiver several replicas of the sameinformation signal transmitted over independently fading

channels, the probability that all the signal components will

fade simultaneously is reduced considerably.

5.1 Maximal Ratio Combining (MRC)In telecommunications, maximal-ratio combining [9] is a

method of diversity combining in which: (a) the signals

from each channel are added together,(b) the gain of each

channel is made proportional to the rms signal level and

inversely proportional to the mean square noise level in that

channel.(c) different proportionality constants are used for

each channel. It is also known as ratio-squared

combining and predetection combining. Maximal-ratio-

combining is the optimum combiner for

independent AWGN channels. In this method, the diversity

branches are weighted for maximum SNR as can be seen in

Figure 5 [9].

The Combiner output is given by

∑ ………..... (12)

The SNR of the combined signal is

∑

….……… (13)

5.2 Equal-Gain Combining Diversity (EGC )

Various techniques are known to combine the signals

from multiple diversity branches. In Equal Gain

Combining,[10] each signal branch weighted with the samefactor, irrespective of the signal amplitude. However, co-

phasing of all signals is needed to avoid signal cancellation.

In figure 5 each branch signal is rotated by , all branchsignals are then added.

The Combiner output is given by

∑ ……….… (14)

The SNR of the combined signal is

Γ = ∑

.....……… (15)

VI. RESULTS

If we set the simulation environment for the OFDM basedwireless modulation, then we get the variable performance

for equalizers as well as for diversity techniques.

The following results have been obtained with the

considered combinations.

Fig. 1.

Fig. 2.

Fig. 3.

0 3 6 9 12 15 18 21 24 27 30

10-0.15

10-0.13

10-0.11

10-0.09

10-0.07

10-0.05

10-0.03

SNR in dB

B i t E r

r o r R a t e

o f E G

C

BER with 'LS', 'MRC', 'ZF', 'EGC' with 512 Subchannels, 32 QAM and 1000 iterations

LS

MRC

ZF

EGC

0 3 6 9 12 15 18 21 24 27 30

10-0.13

10-0.11

10

-0.09

10-0.07

10-0.05

10-0.03

SNR in dB

B i t E r r o

r R

a t e

o f E G

C

BER with 'LS', 'MRC', 'ZF', 'EGC' with 256 Subchannels, 32 QAM and 1000 Iterations

LS

MRC

ZF

EGC

0 3 6 9 12 15 18 21 24 27 30

10-0.16

10-0.14

10-0.12

10-0.1

10-0.08

10-0.06

10-0.04

10-0.02

SNR in dB

B i t E r r o

r R a t e

o f E G

C

BER with 'LS', 'MRC', 'ZF', 'EGC' with 1024 subchannels, 32 QAM and 3000 iterations

LS

MRC

ZF

EGC



164




Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

Fig. 8.

VII. CONCLUSION

In this paper we have considered the maximum number

equals 3000. We have also considered varying number of subchannels. The wireless communication without the

channel estimation results in high errors. Therefore

channel estimation is important to know the parameters of

the channel and also to get the knowledge of affecting parameters. Without the channel estimation we can not

find the proper knowledge of the channel andimpairments.

If we look at the simulation result with 1024

subchannels and 1000 iterations, we find that the BER curve is becoming linear and therefore with high number

of subcarriers and with 1000 iterations we get more better

results compared to 256 and 512 subchannel conditions.Figure no.7, shows the effect of the

different equalizers and diversity techniques with the

2000 iterations and 1024 subchannels. Here we find the

results in the range of 0.1 to 0.001. But not a good

difference is seen here.

One thing is clear here that with lower order QAMmodulation techniques results are not much comparative,

but results in bunching like the optimum performers.Only the change between performances can be seen with

lower number of subcarriers and with 1000 iterations.

Therefore we conclude with this assumption that with

more number of iterations and with the higher number of subchannels results can be improved but at the cast of high

simulation time. Finally with 1000 iterations we have

better performance of used algorithms and MRC is verymuch able to show expected results and with less

complexity in achieving the better BER performance.

ACKNOWLEDGMENT

Foremost, I would like to express my sincere gratitude

to my advisor Prof. Ravi Mohan for the continuous

support of my M.Tech study and research, for his patience,motivation, enthusiasm, and immense knowledge. His

guidance helped me in all the time of research and writing

of this thesis. I could not have imagined having a better advisor and mentor for my M.Tech study.

Besides my advisor, I would like to thank Prof. L. D.

Malviya for their encouragement, insightful comments,

0 3 6 9 12 15 18 21 24 27 30

10-0.04

10-0.03

10-0.02

SNR in dB

B i t E

r r o r R a t e

o f E G

C

BER with 'LS', 'MRC', 'ZF', 'EGC' with 64 subchannels, 32QAM and 1000 iterations

LS

MRC

ZF

EGC

0 3 6 9 12 15 18 21 24 27 3030

10-0.38

10-0.36

10-0.34

10-0.32

10-0.3

10-0.28

10-0.26

10-0.24

SNR in dB

B i t E r

r o r R a t e

o f E G

C

BER with 'LS', 'MRC', 'ZF', 'EGC' with 64 Subchannels, 4 QAM, and 3000 iterations

LS

MRC

ZF

EGC

0 3 6 9 12 15 18 21 24 27 30

10-0.38

10-0.36

10-0.34

10-0.32

10-0.3

10-0.28

10-0.26

10-0.24

SNR in dB

B i t E r r o

r R a t e

o f E G

C

BER with 'LS', 'MRC', 'ZF', 'EGC' with 64 subchannels, 4 QAM and 1000 iterations

LS

MRC

ZF

EGC

0 3 6 9 12 15 18 21 24 27 3010

-3

10-2

10-1

100

SNR in dB

B i t E r r o

r R a t e

o f E G

C


LS

MRC

ZF

EGC

0 3 6 9 12 15 18 21 24 27 3010

-3

10-2

10-1

100

SNR in dB

B i t E

r r o r R a t e

o f E G

C


LS

MRC

ZF

EGC



and hardthank my

spiritually

[1] B. Y

for para

Vol.

[2] Y.LI

wire

chan

sept.

[3] Yus

syste

[4] ww

[5] B.

syste

[6] Theo

and

[7] John

5th e

[8] Hafe

wire

tech

[9] A.

maxi

com

inter

1463

[10] Y. S

for E

Rayl

pp. 8

AUTHO

uestions. Lastfamily &

throughout m

RE

ang.K.B. Letaief,

FDM transmissietric channel

49, no. 3, pp 467-

, J.H. Wintess, a

ess communicat

nel estimation,”IE

2002, pp. 1471–1

i Shen and Ed

ms’” January 200

.google.co.in

. Lathi, “Mode

ms“.

dore S. Rappapo

ractice,” 2nd Edit

G. Proakis and

dition

th Hourani, “A

ess communicat

ology.

hah and A.M.

mal ratio comb

ining for mobil

erence”, IEEE

, Jul 2000.

ng, S. D. Blostei

qual Gain Combi

eigh Fading,” IEE

65-870, Sept. 200

’S PROFILE

DeepmDOB - 3rd I have co

communic

of technol

Master of

Technolog

but not the ly friends

life.

FERENCES

R.S. Cheng, Z. C

n in multipath f odeling,“ IEEE

478, march 2001.

d N.R. Sollenbe

ions, Signal de

EE Trans. Comm

477.

artine Z, “Chan

6.

rn Digital and

rt, “Wireless Co

ion.

asoud Salehi, “

overview of

ion systems,”

Haimovich, “P

ining and com

radio communi

rans. Veh. Tech

and J. Cheng, “

ing EGC With C

E Trans. Wireless

3.

la Singh Pact. 1984 at Satna pleted my grad

tion engineering

ogy & science ,

engineering fro

, Jabalpur.

o ri ht © 2

ast, I would lor supportin

ao,“ Channel esti

ding channels baTrans. Commun

ger “MIMO-OF

tection with en

nication, Vol. 50

el estimation in

analog Commu

munication, Pri

igital Communic

iversity techniq

elsinki Univers

rformance anal

arison with o

cations with coc

ol., vol. 49, pp.

xact Outage Pro

ochannel Interfer

Commun., vol. 2

rihar

ation in electro

from Vindhya i

Satna, Now p

Shri Ram Insti

Internation

012 IJECCE,

ke tome

mation

sed oncation,

M for

hanced

, no. 9,

FDM

ication

nciples

ations,

es in

ity of

sis of

timum

hannel

1454-

ability

nce in

, no. 5,

ics &

stitute

rsuing

tute of

l Journal of Ele

ll ri ht reserv

tronics Commu

V

ed

ication and Co

olume 3, Issue 2,

165

puter Engineeri

ISSN 2249 –071

g

X

ber performance improvement using fading techniques

Documents