TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol.15, No.1, July 201
5, pp
. 120 ~ 127
DOI: 10.115
9
1
/telkomni
ka.
v
15i1.803
7
120
Re
cei
v
ed Ma
rch 2
8
, 2015;
Re
vised Ma
y 9, 2015; Acce
pted May 2
5
, 2015
Symbol Error Rate Performance Analysis of Decode
and Forward Cooperative Communication System
Qabas Ali
Hikmat
1
*, Bin Dai
1
, Rokan
Khaji
2
, Benxi
ong Hua
n
g
1
, Edriss Eisa
1
1
Departme
n
t of Electronics a
n
d
Information E
ngi
neer
in
g
Huaz
hon
g Un
i
v
ersit
y
of Sci
e
n
c
e and T
e
chno
log
y
W
uhan 4
3
0
074
, P.R. China
2
Dep
a
rtment of
Mathemati
cs, Coll
eg
e of Scie
nce
Univers
i
t
y
of Di
ya
la, Di
ya
la
32
001
, Iraq
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: qabas
al
biat
y@
yah
oo.com
A
b
st
r
a
ct
In this pap
er, w
e
investig
ate the p
e
rfor
manc
e
of dec
od
e an
d forw
ard
(DF
)
in multi
p
le-re
la
y
networks. We
consider a c
ooperativ
e divers
ity system
w
here a source s
e
nds infor
m
ation
to the destination
w
i
th the ai
d of
mu
ltirel
ay w
o
rking i
n
DF
re
l
a
yin
g
prot
ocol
and
inv
e
stiga
t
e the o
p
timal
pow
er al
loc
a
ti
on
(OPA) at both
the sourc
e
an
d the re
lay n
o
des. By takin
g
adva
n
tag
e
of
the mo
me
nt g
ener
ating fu
nct
i
on
(MGF
)
, a close
d
for
m
ex
press
i
on for
the sy
mbol
error r
a
te (
SER) an
d si
gn
al to
nois
e
rati
o (SNR) for
bo
th,
M phase shift
keying (MPSK)
and M
quadr
ature amp
litude m
o
dulation
(M
QAM)
signals has been der
iv
ed
to illustrat
e
the asymptotic perform
ance
of the DF system
w
here the ap
proxi
m
ation SE
R is
tight at
hi
gh
SNR. Results
show that th
e
proposed syst
em
,
bas
ed on
SER lower bo
und
is tight to the theoretic
a
l S
E
R
upp
er b
o
u
nd,
and
the
sug
g
e
s
ted OPA
out
perfor
m
s th
e
equ
al
pow
er
a
llocati
on
(EPA
) an
d at
differ
ent
nu
mb
er of rela
ys.
Ke
y
w
ords
: W
i
reless co
mmun
i
catio
n
, Coo
per
ative co
mmu
n
i
c
ations, Dec
o
d
e
and F
o
rw
ard,
Symb
ol error
rate, Optim
a
l po
w
e
r alloc
a
tio
n
.
Copy
right
©
2015 In
stitu
t
e o
f
Ad
van
ced
En
g
i
n
eerin
g and
Scien
ce. All
rig
h
t
s reser
ve
d
.
1. Introduc
tion
Re
cently the best pro
p
o
s
ed techni
que
to design virtual anten
na
array
s
witho
u
t using
collo
cate
d m
u
ltiple a
n
ten
nas is a
c
hie
v
ed by
u
s
in
g coop
erativ
e dive
rsity n
e
tworks. Th
e
s
e
netwo
rks
use
the neig
hbo
r
node
s to
sup
port the
so
ur
ce by se
ndin
g
the so
urce i
n
formatio
n to th
e
destin
a
tion fo
r a
c
hievin
g
spatial dive
rsit
y. The In
vent
ion of
coo
p
e
r
ative com
m
u
n
icatio
ns is
n
o
t
limited only to the physi
cal layer. It is pre
s
ented
in
various fo
rm
s at different
higher p
r
oto
c
ol
layers
[1].
The mo
st co
mmon a
ppli
c
ations
of co
o
perative dive
rsity are
cellul
a
r an
d ad
-h
o
c
wi
rele
ss
comm
uni
cati
on sy
stems [
2
]. The idea
of coo
perativ
e diversity ha
s bee
n recen
t
ly introduced
to
overcome th
e pro
b
lem
of spa
c
e li
mitations in
cellul
a
r a
n
d
ad-ho
c n
e
tworks. Mul
t
i-hop
transmissio
n i
s
a
spe
c
ial
case
of a broa
der
cla
ss
of transmi
ssion
p
r
otocols,
re
ce
ntly which ha
ve
been
re
ceivin
g sig
n
ifica
n
t a
ttention in va
rious
co
mmun
i
ties [3]. In o
r
der to
tran
smi
t
information
i
n
wirel
e
ss ad
-h
oc net
wo
rks, coo
per
ative diversity is a n
o
vel techni
qu
e pro
p
o
s
ed f
o
r a
c
hieving t
h
is
pro
c
e
ss. In [4
] and [5], the authors p
r
o
p
o
se
d seve
ral
topologi
es fo
r coo
perative ad-h
o
c
netwo
rks
to reali
z
ing th
e perfo
rma
n
ce. Some re
se
arche
r
s a
r
e f
o
cu
se
d on m
odifying the i
n
terferen
ce a
nd
noise by applying the ad-hoc c
onfiguration to the cooper
ative nodes [6],[7]. Cooperative
comm
uni
cati
on ha
s b
een
con
s
id
ere
d
as a
goo
d m
e
thod to d
e
velop
comm
u
n
icatio
n qu
ality o
f
s
e
r
v
ic
e (
Q
o
S
)
in a
w
i
r
e
less
ne
tw
ork
s
,
w
i
th
mo
b
ile ad
-
h
oc n
e
t
w
o
rk
s (
M
AN
ETs
)
[8
]. T
h
e a
u
t
ho
r
s
in [9] sugge
sted a topology
which aims t
o
kee
p
the energy path
s
efficient and
decrea
s
e p
o
w
er
con
s
um
ption
in the netwo
rk.
In this pap
er,
we first deriv
ed a cl
osed form
symbol
error rate (S
ER) formulati
on for M
pha
se shift keying (MPSK
) and M qu
a
d
ratu
re ampli
t
ude modul
ation (MQAM
)
signal
s usi
ng the
moment ge
ne
rating fun
c
tio
n
(MGF
) of the received
si
gnal to noise ratio (SNR) at the destinati
on,
sin
c
e the SER formul
ation
is too compli
cated,
we then find a tight lowe
r bou
nd
whi
c
h co
nverges
to the sam
e
li
mit as the th
eoreti
c
al u
p
p
e
r bo
und
i
n
h
i
gh SNR. Co
nse
que
ntly makin
g
u
s
e of t
h
is
result, we develop an opti
m
al po
wer all
o
cation (OPA) method to
minimize the
SER and illustrate
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 2302
-40
4
6
TELKOM
NIKA
Vol. 15, No
. 1, July 2015
: 120 – 127
122
by simul
a
tion
s the
perfo
rm
ance devel
op
ment
of
ou
r new sche
me comp
ared
to the
equa
l p
o
w
e
r
alloc
a
tio
n
(EPA) scheme fo
r different nu
mber
of relays used in the
system.
The
re
st of t
he p
ape
r i
s
orga
nized
as follo
ws. In
section
2,
we
descri
b
e
the
syste
m
model a
nd propo
se a
class of co
ope
rati
on protocols f
o
r multi-node
wirel
e
ss net
works. In se
ction
3, we an
alyze the SER pe
rforma
nce by
using th
e co
nce
p
t of MGF
and obtai
ned
two bou
nd
s for
the exact SE
R exp
r
e
ssi
on
. In se
ction
4
,
we d
e
termi
ne the
OPA for the
tight S
E
R lo
wer bo
und
and
explain
it with t
w
o t
y
pes
of modulation
signals: MPSK and M
Q
AM
modulation.
The
simulatio
n
re
sults
a
r
e pre
s
ented
i
n
se
ction
5. Fi
nally
se
ction
6 the
con
c
lu
sio
n
s d
e
rived f
r
om
the
results a
r
e st
ated.
2. Sy
stem Model
For the DF
strategy, the node
s de
code
each
symbol
of the message an
d tran
smit the
decode
d
symbol ove
r
o
r
th
ogon
al
cha
n
n
e
ls;
relays ap
ply som
e
fo
rm of d
e
codin
g
alg
o
rithm
s
to
their received
sign
als
and
re-en
c
o
de the
informatio
n i
n
to their t
r
an
smitted
sign
a
l
s. In a
wirel
e
ss
coo
perative comm
uni
cati
on system,
sho
w
n in
Fi
gure
1
, the source comm
unicates
with
th
e
destin
a
tion t
h
rou
gh
existing rel
a
ys,
usin
g DF rel
a
ying. All terminals are a
s
sumed
to b
e
sup
p
lied
with singl
e anten
n
a
transmitter
and re
ceive
r
.
.
Figure 1. System model wi
th multi relay (
,
,...,
)
We
sup
p
o
s
e
that the main
cha
nnel
gai
ns a
nd the
chann
el state
informatio
n (CSI) are
kno
w
n
at the
destin
a
tion .
A
ll use
r
s tran
smit si
gnal
s
are th
rou
gh
orthog
onal
chann
els
by u
s
ing
TDMA, FDM
A
or CDMA schem
e.
We divide
sig
nal tran
smi
s
sion into two
pha
se
s. In phase 1, the s
ource b
r
oa
dcasts th
e
sign
al to the
destin
a
tion
and to all
rel
a
y node
s in
the network.
The received
sign
als
at the
destin
a
tion (
D
) an
d at the relay nod
es
(
) in the first available time sl
ot are mod
e
le
d as follo
ws:
(1)
(2)
whe
r
e
is the transmitted i
n
formatio
n sy
mbol from th
e sou
r
ce,
illustrate the
received signal
from nod
e
to node
k
,
are the fading cha
nnel coefficient
s from n
ode
to node
, and
are the co
rre
s
po
ndin
g
add
itive white G
aussia
n
noise (AWG
N) wi
th variance
from nod
e
to
node
.
In pha
se
2, th
e source
and
the rel
a
ys tra
n
smit
sign
al to the d
e
stin
ation. No
w, the
relay
s
corre
c
tly decode the received sig
nal.
Acco
rdin
gl
y, the receive
d
sign
als at
the destinat
ion
terminal are as
follows
:
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Sym
bol Erro
r Rate Perfo
r
m
ance Anal
ysis of Decode
and Forwa
r
d
…
(Qa
b
a
s
Ali Hikm
at)
123
̅
(3
)
whe
r
e
̅
is the decod
ed info
rmation at the
relay
,
and
are the power at the sou
r
ce and
at the
relay resp
ectively. From the
s
e ex
pre
ssi
on
s, we
derive the followin
g
relatio
n
s:
|
|
(4)
(5)
whe
r
e
and
are the insta
n
taneo
us SN
R of the dire
ct and the
r
e
lay r
e
spec
tively. The
variable
is defined befo
r
e, modeled
as the inde
p
ende
nt zero-mean, ci
rcularly-symmetri
c
compl
e
x Gau
ssi
an rando
m
variable
with
varian
ce on
e. By using the maximum
ratio com
b
in
er
(MRC), the
si
gnal
s from
so
urce an
d
relay are
com
b
i
ned at the
de
stination
and
the re
ceived
SNR is:
∑
(6)
γ
and
γ
follow expone
ntial d
i
stributio
n wit
h
para
m
eters:
(
7
)
(8)
whe
r
e
and
are the varia
n
ce of
h
and
h
respec
tively.
3. SER Perfo
r
mance Anal
y
s
is
In this se
ctio
n, we an
alyze the SER p
e
rf
orm
a
n
c
e o
f
multi relay system to ev
aluate
decode
and forward transmissi
on,
and
determi
ne the SER using M
PSK and MQAM signals over
the Rayleig
h
fading
cha
nne
l as follow:
A. MPSK
signals
The co
ndition
al SER with SNR
is descri
bed in [10], for MPSK modulation si
gnal:
,
(9)
By averaging
the conditio
n
a
l SER, repre
s
ente
d
by (9) over the allo
cation of
and
, then the
unconditional
SER for MPSK signal
s of
the proposed
system i
s
given as:
,
∏
(10)
whe
r
e
/
with
⁄
,
and
are the MGF of
and
respe
c
tively.
Substituting the MGF’
s expre
ssi
on, the
n
relation
(10
)
can b
e
give
n as:
,
⁄
⁄
∏
(11)
Equation
(11) represents t
he ex
act expression for the SER fo
r
M
PSK of the proposed
syst
em.
This exp
r
e
ssi
on is not plia
ble in analysi
s
, so we d
e
ri
ve a SER lower bou
nd that
is conve
r
ge
s to
the sam
e
limi
t
as a theo
ret
i
cal SER u
p
p
e
r bo
und to
apply pe
rformance an
alysis.
We set u
p
a
tight SER lower bou
nd u
s
in
g the truth that
0
≤
≤
1
as
following:
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 2302
-40
4
6
TELKOM
NIKA
Vol. 15, No
. 1, July 2015
: 120 – 127
122
⁄
∏
∏
(12)
whe
r
e
.
B. MQAM
sign
als
No
w we
will
derive th
e ex
act SER fo
r
MQAM si
gnal
s. The
conditi
onal SER wit
h
SNR
γ
is written as:
,
(13)
With ave
r
agi
ng eq
uation
(13)
over the
allocation of
and
, then the un
co
nditio
nal SER fo
r
MQAM sig
nal
s of the pro
p
o
s
ed
system i
s
:
,
∏
(14)
whe
r
e
1
1
1
√
⁄
,
/2
with
3
1
⁄
,
and
are
the MGF of
and
respe
c
tively. Substituting the MGF’
s
expressio
n
, then equ
ation (14
)
can
be written a
s
:
,
⁄
⁄
∏
(15)
Equation (15
)
is the exa
c
t expre
ssi
on fo
r
the SER for MQAM of the pro
p
o
s
ed
system.
As in pa
rt A, this expressi
on is
not plia
ble in
an
alysi
s
, so
we
deri
v
e a SER lo
wer
bou
nd th
at is
conve
r
ge
s to
the same li
mit as a th
eo
retical
SER u
pper bou
nd t
o
apply p
e
rfo
r
man
c
e
analy
s
is.
We set up a tight SER lowe
r boun
d usi
n
g
the truth that
0
≤
≤
1
as fo
llowing:
⁄
∏
∏
(16)
whe
r
e
2
, proof
see
s
the app
endix.
Figure 2. The
simulation S
E
R lowe
r bo
u
nd and a
nalytical up
per b
o
und versu
s
SNR for MPSK
sign
al.
0
5
10
15
20
25
30
35
40
10
-10
10
-8
10
-6
10
-4
10
-2
10
0
P/
N
o
(
d
B)
SE
R
A
n
a
l
y
t
i
c
al
U
p
p
e
r
B
o
und
Si
m
u
l
a
ti
o
n
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Sym
bol Erro
r Rate Perfo
r
m
ance Anal
ysis of Decode
and Forwa
r
d
…
(Qa
b
a
s
Ali Hikm
at)
123
Figure 3. The
simulation S
E
R lowe
r bo
u
nd and a
nalytical up
per b
o
und versu
s
SNR for M
Q
A
M
sign
al.
Figure 2 and 3 represent the
SER obtained from ou
r system in rel
a
tion with SNR (
⁄
)
of DF m
u
lti-relay system
with MPSK
and M
Q
AM
signals respectively
. In our
system,
the
varian
ce
s of the cha
nnel coeffi
cient
s are equal to one. Assume:
⋯
and
1
. From the two figure
s
, we can ob
se
rve t
hat the
simulatio
n
of SER lower boun
d and
analytical up
p
e
r bou
nd are very clos
e especi
a
lly in the high SNR re
gime. In this result, the exact
SER expre
ssi
on is bo
und
with two tight boun
ds that i
s
ca
n be con
s
ide
r
ed a
s
im
portant.
4. Optimal Po
w
e
r Allocation
In this
s
e
c
t
ion, we aim to illus
t
rate the as
ymptotic p
e
rforman
c
e of t
he syst
em by
finding
the OPA to th
e tight SER l
o
wer bo
und. T
he mai
n
ide
a
is that
we try
to find the
OPA for the
multi-
node
sy
stem
that minimi
ze
s the
SER l
o
wer bo
und.
B
y
usin
g the
total fixed tran
smissio
n
p
o
wer
denote
d
by
as in [11] to
find the optimal powe
r
at the sou
r
ce
a
nd the
relay
. In
the
following anal
ysis, we
consider the two t
y
pes
of modulation signals:
MPSK and MQAM signal
.
A. MPSK signals
In the propo
sed co
ope
rative system, we
con
s
id
e
r
the
variance of the noi
se is u
n
it, by
s
u
bs
tituting
and
by their values in the lower bou
nd
expression i
n
(12), then we get the
following optimization problem for MPSK modulation
signal:
,
∏
∑
(
17)
After applying
lagran
ge mul
t
iplier app
roa
c
h into form
ul
a (17
)
, and setting
⁄
, then the
next func
tion
is
formed:
,
,
∏
1
∑
(18)
We de
rive the followin
g
function
s by e
m
ploying of the loga
rithm functio
n
in (18
)
:
⁄
∑
0
(19)
⁄
∑
0
(20)
0
5
10
15
20
25
30
35
40
10
-10
10
-8
10
-6
10
-4
10
-2
10
0
P/
N
o
(
d
B)
SE
R
A
n
al
y
t
i
c
al
U
p
per
B
oun
d
Si
m
u
l
a
t
i
o
n
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 2302
-40
4
6
TELKOM
NIKA
Vol. 15, No
. 1, July 2015
: 120 – 127
122
⁄
1
∑
0
(21)
Acco
rdi
ngly we have to solve the form
ation in (19
)
, (20
)
and (21
)
. Assum
e
all relays have th
e
same p
o
wer,
that
⋯
, and verify the equation as:
1
3
1
1
3
1
2
1
0
(22)
B. MQAM signals
In the propo
sed co
ope
rative system, we
con
s
id
e
r
the
variance of the noi
se is u
n
it, by
s
u
bs
tituting
and
by their values in the lower bou
nd
expression i
n
(16), then we get the
followin
g
optimization p
r
o
b
l
em for MQA
M
modulatio
n
signal:
,
∏
∑
(
23)
With applyin
g
lagran
ge mul
t
iplier app
roa
c
h into eq
uati
on (23
)
, then
we get:
,
,
∏
1
∑
(24)
Then with e
m
ploying of the
logarithm fun
c
tion in (2
4)
we de
rive the
following rela
tions:
⁄
∑
0
(25)
⁄
∑
0
(26)
⁄
1
∑
0
(27)
After solving the above rela
tions, then we get:
2
1
6
1
1
6
1
2
1
2
1
0
(28)
Finally, by so
lving equ
ations
(22
)
an
d (28)
by m
a
tla
b
and m
a
ki
ng
use
of (2
1)
and (27
)
,
we
have th
e
optimum
po
wer value
s
with multi
re
lay system
f
o
r b
o
th MP
SK and M
Q
AM
modulatio
n si
gnal
s:
1
⁄
(29)
By s
ubs
tituting the relation
of
,
and
, we have:
⋯
1
⁄
(30)
5. Simulation Resul
t
s
In this
s
e
c
t
ion, we repr
esent the
OPA res
u
lts us
ing MPSK and MQAM modulation
sign
als to aut
henticate the
mathematica
l
terms in
se
ction 4. In bo
th types of modulatio
ns, we
sho
w
the tigh
tness of the analytic
al ex
pre
ssi
on alo
n
g
with the si
mulation
curv
es at hig
h
SNR.
This be
havio
r is in accord
ance with th
e fact
that at adequ
ately high SNR, the obtaine
d lo
wer
and up
per b
o
und
s co
nverg
e
to the same
limit as dec
la
red in the p
r
e
v
ious an
alytical results.
We com
p
a
r
e the SER performa
n
ce of o
u
r prop
osed OPA techniq
ue (
and
that are
substituted
by
equations (29) and (30)
respectively)
with respect to EPA (
/
1
)
at different n
u
mbe
r
of rel
a
ys in relatio
n
to t
he SNR. From figure 4 and
5, it is cle
a
r that
the
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Sym
bol Erro
r Rate Perfo
r
m
ance Anal
ysis of Decode
and Forwa
r
d
…
(Qa
b
a
s
Ali Hikm
at)
123
proposed OP
A obtains better perform
a
nce than
the EPA method due to
the performance going
parall
e
l alon
g
the theoretical SER uppe
r bound at different nu
mbe
r
s of relay
s
.
Acco
rdi
ng to
the obtaine
d
results, we can se
e that the behavio
r of
our p
r
op
osed
OPA is
the sa
me
wh
en we u
s
e o
n
e
relay, but
when
we in
cre
a
se th
e num
b
e
r of relays, t
he pe
rform
a
n
c
e
become
better tha
n
EPA
espe
cially when
we
u
s
e
numbe
r
of re
lays e
qual
3,
and
that i
s
t
he
rea
s
on that p
r
ompt u
s
to sugge
st use
of multi-relay in
our propo
se
d model.
Figure 4.
SER com
p
arison between suggested
OPA
and EPA for MPSK modulation.
Figure 5. SER com
p
arison between sugge
sted OPA
and EPA for MQAM modul
ation.
6. Conclusio
n
This pa
per p
r
ese
n
ts
a
stru
cture
for e
n
h
anc
i
ng th
e S
E
R p
e
rfo
r
ma
nce
of
DF
m
u
lti-rel
a
y
coo
perative tran
smi
ssi
on
over Rayleig
h
f
ading
ch
annel
s. We
demon
strate
that the SER
perfo
rman
ce
can
be
si
gnif
i
cantly imp
r
o
v
ed by
the
p
r
ope
r
relay
strategy. We
derive
a
clo
s
ed
expressi
on for the SER, by usi
ng the
concept of MGF for M
PSK and MQAM modulation
signals
and compa
r
e
it with theore
t
ical SER up
per bo
und. A
n
OPA sch
e
m
e is inve
stigated to mini
mize
the SER. It can be
seen f
r
om the
simu
lation re
sult
s how o
u
r p
r
o
posed OPA
outperfo
rm
s the
EPA at differ
ent number of relays
.
10
15
20
25
30
35
40
10
-30
10
-25
10
-20
10
-15
10
-10
10
-5
10
0
P/
N
o
(
d
B)
SE
R
OP
A
C
a
l
c
ul
a
t
ed
(
n
=
1
)
EP
A(
n
=
1
)
OP
A
C
a
l
c
ul
a
t
ed
(
n
=
2
)
EP
A(
n
=
2
)
O
P
A
a
pprox
i
m
at
i
o
n
t
h
e
o
ry
(
n
=
2
)
OP
A
C
a
l
c
ul
a
t
ed
(
n
=
3
)
EP
A(
n
=
3
)
O
P
A
a
pprox
i
m
at
i
o
n
t
h
e
o
ry
(
n
=
3
)
10
15
20
25
30
35
40
10
-30
10
-25
10
-20
10
-15
10
-10
10
-5
10
0
P/
N
o
(
d
B)
SE
R
OP
A
C
a
l
c
ul
a
t
ed
(
n
=
1
)
EP
A(
n
=
1
)
OP
A
C
a
l
c
ul
a
t
ed
(
n
=
2
)
EP
A(
n
=
2
)
O
P
A
a
pprox
i
m
at
i
o
n
t
h
e
o
ry
(
n
=
2
)
OP
A
C
a
l
c
ul
a
t
ed
(
n
=
3
)
EP
A(
n
=
3
)
O
P
A
a
pprox
i
m
at
i
o
n
t
h
e
o
ry
(
n
=
3
)
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 2302
-40
4
6
TELKOM
NIKA
Vol. 15, No
. 1, July 2015
: 120 – 127
122
Appe
ndix: p
r
oof of r
e
lati
on (16
)
As we have
0
≤
≤
1
, we ca
n derive the followin
g
ineq
ualities for M
Q
AM sign
als:
0
≤
2
≤
2
2
2
(31)
∏
∏
∏
(32)
Multiplying (3
1) and
(32
)
, then we obtai
n:
∏
∏
∏
(33)
No
w su
bsequ
ently we take
the integral of
(33),
then
we
prove the rel
a
tion (16
)
. We can u
s
e the
same a
s
the
above proced
ure for MPSK
signal
s.
Referen
ces
[1]
Korakis T
.,
T
a
o
Z., Slutskiy
Y., Pan
w
ar S.
A
coop
erative M
A
C protoc
ol
for
ad hoc w
i
re
les
s
netw
o
rks
.
In Pervasive C
o
mputi
ng a
nd
Commun
i
cati
o
n
s W
o
r
kshops
(PerCom W
o
rk
shops). 20
07;
5: 532-5
36.
[2]
Gupta K., Ghosh P. K. End-to
-End Perform
a
nce
of Mu
ltipl
e
-
I
nput-Multi
p
le-
O
utput Rel
a
y T
r
ansmissi
o
n
Link ov
er Ra
yl
eig
h
F
adi
ng C
han
nels.
Intern
ation
a
l Jo
urna
l
of
Computer
Appl
icatio
ns
, (201
2);
53(2)
:
7-12.
[3]
Lan
eman J. N
.
Cooper
ative
communi
c
a
tio
n
s
in mobi
le a
d
hoc net
w
o
rks.
IEEE Signal
Processi
n
g
Maga
z
i
ne
. 20
06
.
[4]
Neves T
.
F
.
, Bordim J. L.
T
opolo
g
y
c
ont
rol i
n
cooper
ative a
d
hoc
w
i
re
less
net
w
o
rks.
Elec
tronic Notes
in T
heor
etical
Co
mp
uter Scie
nce
. 201
4: 302
: 29-51.
[5]
Aish
w
a
r
y
a
S., Hemal
a
tha
R., Yamun
a
D.
M., V
ijitha J.
Desig
n
a
nd A
nal
ysis
of impr
ovin
g Net
w
o
r
k
capac
it
y
w
i
t
h
coop
erative co
mmunicati
ons.
Internatio
na
l Journ
a
l of Adva
nced R
e
searc
h
in Electrical
,
Electron
ics an
d Instrumentati
on Eng
i
n
eeri
n
g
. 2013; 2(3).
[6]
Gu W., Peters G., Clavi
er
L., Septier F., Nevat I.
Receiv
er
study for c
oop
erative c
o
mmu
n
icati
ons
i
n
co
n
v
ol
ve
d add
i
t
i
v
e
α
-stab
l
e
interfer
ence
plus
Gauss
i
an th
er
mal
nois
e
. IEEE
International
S
y
mp
osi
u
m. 2012; 45
1-4
55.
[7]
Che
n
J., Clavie
r
L., Xi Y
., Burr A., Rolland N.,
Rolla
nd P. A.
α
-stable interfe
r
ence
mo
del
lin
g and re
la
y
selection for
regenerativ
e
c
ooperative IR-
U
WB system
s
.
IEEE Europe
an Wirel
e
ss T
e
chnolog
y
Confer
ence (E
uW
IT
). 2010; pp. 81-84.
[8]
Yu F
.
R.,
T
ang H., Bu S., Z
h
eng D. Secur
i
ty an
d qu
alit
y o
f
service (QoS)
co-desi
gn in
coop
erati
v
e
mobil
e
ad
hoc
net
w
o
rks.
EUR
ASIP Journ
a
l
o
n
W
i
rel
e
ss
Co
mmu
n
icati
ons
and
Netw
orki
n
g
. 2013;
(1):
1-14.
[9]
Pandey
K. K., Baghel N., Sin
ha, S. K. Opti
mum Rel
a
y
S
e
lectio
n
for Ene
r
g
y
-Effici
ent C
oop
erative A
d
Hoc Net
w
orks.
Internation
a
l
Journ
a
l of Mo
dern En
gin
eer
i
ng Res
earch (
I
JMER)
. 2013;
3(2):108
2-
108
7.
[10]
Simon M. K., Aloui
ni M. S. Dig
ital commu
nica
tion over fad
i
n
g
chan
ne
ls: John W
ile
y & So
ns. 2005; 9
5
.
[11]
Goldsmith A. W
i
reless com
m
unic
a
tions: C
a
mbrid
ge U
n
iv
ersit
y
pr
ess: 20
05.
Evaluation Warning : The document was created with Spire.PDF for Python.