Indonesi
an
Journa
l
of El
ect
ri
cal Engineer
ing
an
d
Comp
ut
er
Scie
nce
Vo
l.
1
3
,
No.
2
,
Febr
uar
y
201
9
, pp.
5
34
~
542
IS
S
N: 25
02
-
4752, DO
I: 10
.11
591/ijeecs
.v1
3
.i
2
.pp
534
-
542
534
Journ
al h
om
e
pag
e
:
http:
//
ia
es
core.c
om/j
ourn
als/i
ndex.
ph
p/ij
eecs
User sel
ection p
rot
ocol in
DF c
oo
p
erative n
etworks
with hyb
rid
TSR
-
PSR pr
otocol ba
s
ed fu
ll
-
du
plex e
n
ergy har
ves
ting ov
er
rayleigh
fading c
hann
el: system p
er
f
orma
nce anal
ysis
Phu
Tr
an
Ti
n
1
, Le Anh
Vu
2
,
Tan
N. N
guy
en
3
, Th
an
h
-
L
ong
Nguy
e
n
4
1
Facul
t
y
of Elect
ronic
s T
ec
hno
lo
g
y
,
Industri
al Unive
rsit
y
of
Ho C
hi
Minh C
ity
,
Vi
et
nam.
2
Optoel
e
ct
ron
i
cs
Resea
r
ch
Group
,
Facu
lty
of Elec
tri
c
al
and El
e
ct
r
onic
s E
ng
ine
e
rin
g,
Ton
Duc Thang
Univer
sit
y
,
Vie
t
nam
3
W
ire
le
ss
Com
muni
cations
Rese
a
rch
Group,
Facult
y
of El
ec
tr
ic
a
l &
E
le
c
tronics E
ngine
er
ing,
Ton
Duc Thang
Univer
sit
y
,
Vie
t
nam
4
Cent
er
for
Infor
m
at
ion
T
ec
hnolo
g
y
,
Ho Chi
Min
h
Cit
y
Un
ive
rsit
y
of
Food Indust
r
y
,
Vie
tna
m
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
ul
1
2
, 2
018
Re
vised
Oct
2
0,
2018
Accepte
d
Nov
2,
2018
Coopera
t
ive
co
m
m
unic
at
ion
h
as
bee
n
recen
tly
proposed
i
n
wire
l
ess
comm
unic
at
ion
s
y
ste
m
s
for
exploring
th
e
inhe
r
e
nt
spat
ia
l
d
ive
rs
ity
in
r
ela
y
cha
nne
ls.
In
thi
s
work,
we
inv
est
iga
t
e
the
s
y
st
em
per
form
anc
e
of
the
en
er
g
y
har
vesti
ng
ful
l
-
d
uple
x
(FD
)
de
co
de
-
and
-
forward
(
DF
)
h
y
br
id
ti
m
e
sw
it
chi
ng
-
power
split
ti
ng
r
el
a
y
i
ng
TSR
-
PSR
(TPSR
)
proto
col
r
ela
y
ing
netw
ork.
In
th
e
sele
c
ti
on
sch
eme,
th
e
best
user
se
l
ec
t
ion
proto
col
i
s
proposed
and
i
nvesti
gated.
Mainly
w
e
d
erive
the
cl
os
ed
-
f
orm
expr
ession
for
the
out
age
proba
b
il
i
t
y
,
s
y
stem
throughp
ut
and
th
e
s
y
m
b
ol
err
or
r
at
e
(SE
R)
of
the
s
y
stem
.
Num
eri
cal
res
ult
s
are
al
so
pre
sen
te
d
b
y
t
he
Mont
e
C
arl
o
sim
ula
ti
on
to
val
id
at
e
th
e
the
ore
ti
c
al
an
aly
sis
in
connect
i
on
with
th
e
al
l
poss
ibl
e
par
amete
rs
in
th
e
compari
son
be
t
wee
n
TSP
R,
TS
R
and
PS
R
ca
se
s.
Th
e
r
ese
ar
ch
result
s
show
tha
t
TPS
R
c
ase
i
s
bet
t
er
tha
n
the
othe
rs
in
t
erm
of
outa
ge
proba
b
il
i
t
y
and
SER
.
Ke
yw
or
d
s
:
Decode
-
an
d
-
f
orwa
rd (DF
)
Energy
harvest
ing
(E
H)
Fu
ll
-
duplex
(
F
D)
Re
la
yi
ng
n
et
w
ork
User sel
ect
ion
Copyright
©
201
9
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
Le A
nh
Vu,
Op
t
oelect
ronic
s Resea
rc
h Gro
up
Faculty
of Elec
tric
al
an
d El
ect
ronics E
nginee
rin
g
To
n Du
c
Th
a
ng
Un
i
ver
sit
y
Ho Chi Mi
nh
Ci
ty
, V
ie
tnam
.
Em
a
il
:
le
anh
vu@tdtu
.edu.
vn
1.
INTROD
U
CTION
In
recent
ye
ar
s,
the
wi
reless
transm
issi
on
has
ex
pe
rience
d
ra
pid
dev
el
opm
ent.
Re
la
yin
g
has
been
pro
ved
to
be
a
n
e
ff
ic
ie
nt
wa
y
to
exte
nd
th
e
co
ver
a
ge
a
re
a
of
wireless
netw
orks
a
nd
i
ncr
ease
tra
ns
m
issi
on
reli
abili
ty
with
ou
t
a
dd
it
io
nal
transm
it
power
at
t
he
tran
sm
i
tt
ers.
H
ow
e
ver,
th
e
for
wardin
g
operati
ons
at
the
relay
sti
l
l
req
ui
re
e
xtra
e
nergy
.
F
or
relay
no
de
s
powe
re
d
by
powe
r
li
m
it
ed
batte
ries
instea
d
of
t
he
regular
powe
r
gr
i
d,
s
uch
as
m
ob
il
e
de
vices,
s
uch
e
xtra
e
nerg
y
con
s
um
ption
m
ay
cause
seriou
s
c
oncer
ns
.
Re
centl
y,
harv
est
ing
energy
f
ro
m
am
bient
rad
io
-
frequ
e
ncy
(R
F)
en
er
gy
wa
s
pr
opos
e
d,
a
nd
it
has
been
well
stud
ie
d.
T
he
id
ea
that
receivin
g
in
for
m
at
ion
a
nd
ha
rv
est
in
g
en
er
gy
sim
ultaneousl
y
is
not
onl
y
ap
peali
ng
but
al
s
o
w
or
th
furthe
r
inv
est
igati
ng.
The
rece
nt
wor
ks
s
ug
gested
that
data
tra
nsm
itti
ng
a
nd
batte
ry
cha
rg
i
ng
c
ou
l
d
be
f
ulfill
ed
at
t
he
sam
e tim
e, and
the
refor
e
,
the
com
bin
ing
of e
nergy
ha
rv
est
i
ng m
odule a
nd r
el
ay
could
be
a
n
a
vaila
ble
s
olu
ti
on
of extra
en
e
r
gy cons
um
ption
pro
blem
at the relay
[
1
-
6].
Re
centl
y,
de
pl
oying
e
nergy
ha
rv
est
in
g
(E
H)
relay
s
that
util
iz
e
the
e
nergy
colle
ct
ed
from
the
sourc
e
sign
al
f
or
data
tra
ns
fe
r
has
r
ecei
ved
c
on
si
de
rab
le
at
te
ntio
n.
W
it
h
E
H
c
apab
il
it
ie
s,
relay
s
can
be
ins
ta
ll
ed
conve
niently
w
it
ho
ut
wi
rin
g
c
os
t
a
nd
t
he
nee
d
f
or
batte
ry
re
placem
ent.
Be
sides,
substa
ntial
transm
issi
on
powe
r
can
be
sa
ve
d
be
cause
the
inte
r
-
node
distance
is
sh
ort
ene
d
usi
ng
m
ulti
ho
p
tr
ansm
issi
on
s.
H
ow
e
ve
r,
in
form
at
io
n
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
User
sel
ect
io
n pro
t
oc
ol in DF
coop
e
ra
ti
ve
ne
t
works wi
th
hy
br
id
T
SR
-
P
SR
p
r
oto
c
ol
ba
se
d… (Ph
u Tra
n Ti
n
)
535
relay
ing
c
ons
um
es
add
it
io
nal
res
ources
(i.e.
,
ti
m
e
and
ba
ndwi
dth)
c
om
par
ed
to
t
he
dire
ct
transm
issi
on
an
d
hen
ce
so
m
e
early
wo
r
k
has
be
en
c
onduct
ed
f
ocu
sin
g
on
ass
essing
t
he
feasibil
it
y
of
EH
rel
ay
s.
The
gain
offe
re
d
by
the
E
H
rela
y
base
d
on
tim
e
s
witc
hing
rel
ay
ing
(TSR
)
is
analy
ze
d
t
heoreti
cal
ly
in
[
7]
for
both
de
co
de
-
an
d
-
forw
a
r
d
(
DF
)
a
nd
am
plify
-
an
d
-
forwar
d
(
AF)
relay
ing
a
ppr
oach
e
s
with
a
s
ign
ific
a
nt
ga
in
.
The
E
H
relay
base
d
on
DF
is
al
so
consi
der
e
d
in
[
8],
w
her
e
the
r
el
ay
determ
ines
wh
et
her
to
pe
rfor
m
EH
or
i
nfor
m
at
ion
rel
ay
ing
befor
e
t
he
sou
rce
tran
sm
issi
o
n
acco
r
ding
to
a
gr
ee
dy
poli
cy
.
On
t
he
ot
he
r
ha
nd,
an
a
na
ly
ti
cal
fr
am
e
work
is
pro
po
se
d
i
n
[9]
to
e
val
uate
the
perform
ance
of
the
EH
re
la
y
based
on
AF
for
both
T
SR
an
d
powe
r
sp
li
tt
in
g
relay
ing
(PSR
)
unde
r
the
ef
fe
ct
of
the
crit
ic
al
syst
e
m
par
a
m
et
ers
su
c
h
a
s
noise
var
ia
nc
es,
s
ource
t
o
relay
distance,
tra
nsm
issi
on
rate,
a
nd
e
nergy
c
on
ver
si
on
e
ff
ic
ie
ncy.
M
o
re
over
,
relay
sel
ect
io
n
(RS
)
is
a
pr
act
ic
al
appr
oach
t
o
ba
la
nce
the
tra
deoff
betwee
n
reli
abili
ty
i
mp
r
ovem
ent
and
sp
e
ct
ral
eff
i
ci
ency
loss
due
t
o
inf
or
m
at
ion
re
petit
ion
[
10
]
.
I
n
[
11
]
,
tw
o
R
S
sc
hem
es
ai
m
ing
to
at
ta
in
the
op
ti
m
al
tradeoff
betwee
n
ene
r
gy
trans
fer
a
nd
outa
ge
pr
ob
a
bili
ty
/e
rg
odic
ca
pa
ci
ty
f
or
DF rel
ay
s ar
e st
ud
ie
d.
In
this
work,
w
e
pr
opos
e
a
nd
i
nv
e
sti
gate
the
s
yst
e
m
per
f
orm
ance
of
t
he
ene
rg
y
ha
rv
e
sti
ng
fu
ll
-
duple
x
(F
D
)
dec
od
e
-
a
nd
-
f
orwa
rd
(
D
F)
hybri
d
T
SR
-
PSR
protoc
ol
relay
ing
netw
ork
.
F
or
detai
ls
o
n
this
a
naly
sis,
the
energy,
a
nd
in
f
or
m
at
ion
are
tr
ansf
e
rr
e
d
from
the
s
ource
to
the
relay
nodes
,
an
d
al
l
cha
nn
e
ls
are
c
on
si
dered
a
s
the Ray
le
igh
fa
ding c
hannels.
The
m
ai
n
co
ntr
ibu
ti
ons
of the
pap
e
r
a
re sum
m
arized as
fo
ll
ow
s:
1.
The
syst
em
mo
del
of
the
e
ne
rg
y
ha
r
vestin
g
fu
ll
-
duplex
(FD)
decode
-
an
d
-
f
orward
(
DF)
hybri
d
TSR
-
P
SR
protoc
ol
relay
ing
netw
ork
over
t
he
Ra
yl
ei
gh
fa
ding
c
hannel
s
a
nd
t
he
c
om
par
iso
n
bet
ween
hy
br
i
d
TSR
-
P
SR
(TPS
R
),
TSR a
nd PS
R
cases
a
re
propose
d
a
nd in
vestigat
e
d.
2.
The
cl
os
e
d
-
for
m
exp
ressio
ns
of
the
outa
ge
pro
ba
bili
ty
and
the
syst
em
thro
ug
hput
a
re
de
rive
d.
M
or
e
ov
er,
the b
e
st use
r
se
le
ct
ion
pr
oto
c
ol
is pro
posed
a
nd in
vestigat
ed
.
3.
The
sym
bo
l Er
ror
Ra
ti
o (SER
)
a
naly
sis of th
e pro
posed
m
od
el
syst
em
is p
resen
te
d
a
nd de
m
on
strat
ed
.
4.
The
in
flue
nce
of
the
m
ai
n
pa
ram
et
ers
on
t
he
syst
em
per
f
orm
ance
is
dem
on
st
rated
e
ntirel
y
by
t
he
M
onte
Ca
rlo
sim
ulatio
n.
The
struct
ur
e
of
this
pa
pe
r
is
pro
posed
as
f
ollow
s
.
Sect
ions
I
I
pr
ese
nts
t
he
s
yst
e
m
m
od
el
of
the
relay
ing
netw
ork.
Se
ct
ion
s
III
der
ive
s
the syst
e
m
p
erfor
m
ance
of
t
he
m
od
el
syst
e
m
. S
ect
ion
IV
prov
i
des
the
nu
m
erical
resu
lt
s a
nd s
om
e d
isc
us
si
on
s
. F
inall
y, Sect
i
on V co
nclu
de
s the
pa
per
.
2.
SY
STE
M MO
DEL
In
t
his
pap
e
r,
t
he
syst
em
m
od
el
is
the
ene
r
gy
harvesti
ng
f
ul
l
-
duplex
(FD
)
dec
od
e
-
a
nd
-
f
orwa
rd
(
DF)
hybri
d TSR
-
PS
R
pro
toc
ol
relay
ing
net
work
a
s
show
n i
n Fi
gure
1.
I
n t
his
m
od
el
, t
he i
nfo
r
m
at
ion
is
trans
ferred
from
the
s
ourc
e
(
S)
to
the
m
ulti
-
destinat
io
n
(
D
i
),
th
r
ough
e
ne
rg
y
co
ns
trai
ne
d
i
nterm
ediat
e
relay
(R)
.
The
energy
harvesti
ng
a
nd
inf
or
m
at
ion
proces
sin
g
of
th
e
syst
e
m
m
od
e
l
with
hybri
d
TSR
-
PSR
pr
oto
col
are
pro
po
sed
i
n
Figure
2.
In
th
is
schem
e,
T
is
the
blo
c
k
ti
m
e
in
w
hich
the
source
f
ully
transm
it
s
the
info
rm
at
ion
data
to
the
m
ul
ti
-
destinat
ion.
In
the
fi
rst
interval
tim
e
(αT),
t
he
relay
harvests
e
ne
rgy
from
the
s
ource
s
i
gn
al
,
wh
e
re
α
i
s
the ti
m
e
switc
hin
g
facto
r
α
∈
(0,
1)
.
In
the
re
m
ai
nin
g
i
nter
va
l t
i
m
e
(1
-
α
)T
, t
he
relay
node
harvests ρP
s
en
erg
y
from
the
sourc
e
an
d
us
e
(
1
-
ρ)
P
s
ene
rg
y
t
o
tr
ansf
e
r
i
nfor
m
at
ion
to
the
m
ul
ti
-
destinat
io
n
node
s,
w
hich
ρ
is
the
powe
r
sp
li
tt
ing
facto
r
ρ
∈
(0,
1).
All
the
fad
i
ng
c
ha
nnel
s
from
S
to
R
and
R
to
D
a
re
pr
opose
d
as
the
R
ay
le
igh
fad
i
ng
cha
nnel
s.
Mo
re
detai
ls
of
the
a
naly
ti
cal
m
at
he
m
at
ic
a
l
m
od
el
of
t
he
ou
ta
ge
pro
bab
i
li
ty
and
th
rou
ghput
of
t
he
syst
em
m
od
el
is
pr
ese
nted
a
nd
a
naly
zed
in
the
fo
ll
owin
g
sect
io
ns
[
12
-
16
]
.
If
α
=
0,
this
sc
hem
e
beco
m
es
PSP. I
f
ρ =
0
t
hen it
b
ec
om
es
the T
SP pr
otoc
ol.
E
n
e
r
gy H
ar
ve
s
t
i
n
g
(
EH
)
I
n
f
or
m
at
ion
t
r
an
s
m
is
s
ion
(
IT
)
SR
h
S
R
D
1
D
n
D
M
n
RD
h
RR
h
Figure
1. Syst
em
m
od
el
EH
at
R
(
ρ
P
S
)
EH
at
R
α
T
IT
S
R
D
n
(
1
-
ρ
)
P
S
(
1
-
α
)
T
T
Figure
2. The
powe
r
s
plit
ti
ng
pro
t
oco
l
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
2
,
Fe
bru
ary
201
9
:
5
34
–
542
536
3.
THE
SYSTE
M
PE
RFO
R
MAN
CE
In
this
sect
io
n,
we
analy
ze
d
an
d
dem
on
strat
ed
t
he
syst
e
m
per
f
orm
ance
analy
sis
of
the
m
od
el
syst
e
m
[
12
-
14]
.
3.1.
Ener
gy
ha
r
vest
in
g ph
as
e
I
n t
he first i
nte
rv
al
,
the a
ve
ra
ge
tra
ns
m
it
te
d
powe
r
at
the
r
e
la
y ca
n
be
calc
ulate
d
as
:
2
2
1
1
(
1
)
(
1
)
1
s
S
R
r
s
S
R
TP
h
E
P
P
h
TT
(
1
)
wh
e
re
01
is
ener
gy
conver
si
on
e
ff
ic
ie
ncy,
01
is
tim
e
switc
hin
g
f
act
or
,
P
s
is
the
transm
itted
powe
r
at
source
,
h
SR
i
s the so
ur
ce
to rel
ay
ch
a
nn
el
ga
in, respecti
vely
.
In the sec
ond i
nter
val,
on
ce
a
gain
a
ve
rag
e
tr
ansm
itted power at t
he
r
el
ay
can
be give
n
as
:
2
2
2
2
(
1
)
(
1
)
(
1
)
s
S
R
r
s
S
R
T
P
h
E
P
P
h
TT
(
2
)
wh
e
re
01
is t
he
po
wer split
ti
ng
fa
ct
or
.
Finall
y, the am
ount
of ave
rage t
ran
sm
it
te
d
powe
r
at
t
he
rel
ay
can
be ob
ta
i
ned
as
:
2
12
r
r
r
s
S
R
P
P
P
P
h
(
3
)
wh
e
re
we de
note
1
.
3.2.
Tr
an
smi
ssion ph
as
e
T
he recei
ve
d
si
gn
al
at t
he rel
ay
can be e
xpre
ssed
as
:
1
r
S
R
s
R
R
r
r
y
h
x
h
x
n
(
4
)
w
he
re
we
de
note
2
ss
xP
,
2
rr
xP
,
is
ex
pec
ta
ti
on
operat
or,
h
RR
is
lo
opba
ck
inter
fere
nc
e
channel, a
nd
n
r
is t
he
a
ddit
ive
wh
it
e
Gaussi
an
noise (
A
WGN
)
wit
h varianc
e
N
0
.
The recei
ve
d
si
gn
al
at t
he
n
rd
destinat
io
n
ca
n
be f
or
m
ulate
d
as
:
n
n
n
d
R
D
r
d
y
h
x
n
(
5
)
wh
e
re
h
RDn
is
the
relay
to
the
n
rd
de
sti
nation
channel
gain,
n
dn
is
the
a
ddit
ive
w
hite
Ga
ussi
an
noise
(AWGN
)
with
var
ia
nce
N
0
, a
nd
1
,
2
,
.
.
.
,
nM
.
In
t
his
m
od
el
,
we
co
ns
i
der
de
cod
e
a
nd
f
orwa
rd
protoc
ol
(
D
F)
.
F
r
om
(4
),
th
e
sign
a
l
to
noise
rati
o
(SNR
)
at
the r
el
ay
ca
n
be
cal
c
ulate
d
as
2
1
2
0
(
1
)
S
R
s
R
R
r
hP
h
P
N
(
6
)
Substi
tuti
ng (3
)
int
o
(
4) a
nd usi
ng the
fact t
ha
t N
0
<<P
s
, we
hav
e
2
1
2
2
2
0
(
1
)
(
1
)
S
R
s
s
S
R
R
R
R
R
hP
P
h
h
N
h
(
7
)
Fr
om
(
5) a
nd s
ub
sti
tuti
ng
(3), t
he
S
NR at t
he
d
est
inati
on ca
n
be
cal
culat
e
d
as
;
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
User
sel
ect
io
n pro
t
oc
ol in DF
coop
e
ra
ti
ve
ne
t
works wi
th
hy
br
id
T
SR
-
P
SR
p
r
oto
c
ol
ba
se
d… (Ph
u Tra
n Ti
n
)
537
22
2
2
00
nn
r
R
D
s
S
R
R
D
P
h
P
h
h
NN
(
8
)
Finall
y, the e
nd to
end
SN
R
of the
pro
pose
d
syst
em
can
be
obtai
n
as
:
2
1
2
m
i
n
,
ee
(
9
)
In this a
naly
sis, p
le
ase
note th
at
all
o
f
t
he
c
ha
nn
el
belo
ng
t
o
Ra
yl
ei
gh f
a
di
ng
c
ha
nn
el
s
.
Re
ma
r
k 1:
Th
e best user
selec
tion pr
otocol
.
Fr
om
(
9),
we
pro
pose
t
he
opti
m
al
us
er
sel
ect
ion
protoc
ol
i
n
w
hic
h
t
he
best
sel
ect
ion
us
e
r
is
sel
ect
ed
as foll
ows:
2
2
1
,
2
,
.
.
.
,
m
a
x
(
)
n
RD
nM
h
(
10
)
In [17
]
,
the C
um
ula
ti
ve
Di
stri
bu
ti
on
Functi
on
(CDF)
of
2
can
be give
n
by the
fo
ll
owin
g
2
2
/
0
(
)
(
1
)
M
py
pp
M
p
F
y
C
e
(
11
)
w
he
re
2
is t
he
m
ean
of RV
2
,
an
d
!
!
(
)
!
p
M
M
C
p
M
p
.
The
n,
t
he
c
orre
sp
on
ding
Proba
bili
ty
D
ensity
Functi
on
(PD
F)
ca
n
be o
btained
by t
he foll
ow
i
ng
2
2
1
(
1
)
/
1
0
2
1
(
)
1
M
p
py
p
M
p
f
y
C
M
e
(
12
)
Re
ma
r
k 2:
Ou
tage
proba
bil
ity
(
O
P).
Fr
om
(
9),
the
OP
of
DF
syst
e
m
can
be
ex
pre
ssed
as
:
2
1
2
2
2
1
2
0
0
1
2
P
r
P
r
m
in
,
m
a
x
(
1
)
P
r
m
in
,
(
1
)
P
r
m
in
,
n
e
e
t
h
t
h
s
SR
R
D
nM
th
RR
th
OP
P
h
h
N
h
(
13
)
wh
e
re
we de
note
2
22
2
1
2
0
1
0
,
,
m
a
x
,
,
2
1
n
R
s
R
R
S
R
R
D
t
h
nM
P
h
h
h
N
, and
R
is th
e sour
ce
r
at
e.
Fr
om
(
13
),
t
he OP ca
n
be re
w
ritt
en
as th
e f
ol
lowing
0
1
2
(
1
)
1
P
r
P
r
t
h
t
h
OP
(
14
)
In (14),
w
e
de
note
(
1
)
1
(
1
)
(
1
)
Pr
Pr
1
th
th
th
Ie
(
15
)
wh
e
re
is t
he
m
ean
of
RV
.
And
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
2
,
Fe
bru
ary
201
9
:
5
34
–
542
538
12
2
0
1
2
1
2
2
2
0
2
0
2
0
P
r
1
P
r
1
|
(
)
t
h
t
h
th
I
F
f
d
(
16
)
Substi
tuti
ng (1
2) into
(16),
we ha
ve
0
2
1
22
0
2
1
22
1
(
1
)
/
2
1
2
0
2
0
1
(
1
)
/
1
2
0
2
0
1
1
(
1
)
th
th
M
p
p
p
M
p
pp
M
p
M
p
I
C
M
e
e
d
CM
e
e
d
(
17
)
wh
e
re
1
is t
he
m
ean
of
RV
1
.
Applyi
ng table
of inte
gr
al
e
q (3.3
24,
1)
i
n
[18],
the
(
17)
ca
n
be ref
orm
ulated
as
1
2
1
1
0
0
1
2
0
1
2
(
1
)
2
(
1
)
2
(
1
)
M
pp
t
h
t
h
M
p
p
I
C
M
K
p
(
18
)
wh
e
re
()
v
K
is t
he
m
od
i
fied
Be
ssel
functi
on
of
t
h
e
seco
nd k
i
nd and
v
th
or
der
.
Substi
tuti
ng (1
6), (1
8) into
(1
4), OP ca
n
be
c
al
culat
ed
as
(
1
)
1
11
0
0
1
2
0
1
2
1
11
0
0
1
2
0
1
2
(
1
)
1
0
1
2
(
1
)
1
1
2
(
1
)
2
(
1
)
(
1
)
1
2
(
1
)
2
(
1
)
2
(
1
)
(
th
th
M
pp
th
th
M
p
M
pp
th
th
M
p
pp
th
M
p
O
P
e
C
M
K
p
p
C
M
K
p
C
M
e
1
1
0
0
1
2
(
1
)
2
1)
M
th
p
p
K
p
(
19
)
Re
ma
r
k 3:
Th
roughp
ut
.
(
1
)
(
1
)
(
1
)
(
1
)
RT
O
P
O
P
R
T
(
20
)
Re
ma
r
k 4:
Th
e sy
m
bo
l E
rr
or
Rat
i
o
(
SER
) an
aly
sis
In
this
sect
io
n,
we
obta
in
ne
w
ex
pr
e
ssio
ns
for
the
sym
bo
l
Error
Ra
ti
o
(SER
)
at
the
dest
inati
on
.
We
first c
on
si
der the
ou
ta
ge pr
ob
abili
ty
, w
hic
h
was ob
ta
i
ned
in
[19,2
1]. T
hus,
we
have
2
(2
ee
S
E
R
Q
(
21
)
2
/2
1
()
2
x
t
Q
t
e
dx
is
the
Ga
us
si
an
Q
-
functi
on,
a
nd
are
c
onsta
nts
w
hic
h
i
s
sp
eci
fic
f
or
m
od
ulati
on
ty
pe.
(
,
)
(
1
,1
)
for
BP
SK
and
(
,
)
(
1
,
2
)
for
Q
PS
K
.
As
a
re
su
lt
,
be
fore
ob
ta
ini
ng
the
SER
perform
ance,
the
distri
bu
ti
on
functi
on
of
2
ee
is
exp
ect
e
d.
T
he
n,
we
beg
i
n
re
w
riti
ng
the
SER
expressi
on
give
n
in (2
1)
directl
y
in term
s o
f
out
age
pro
bab
il
it
y at
the s
ource
by
u
sin
g
i
ntegr
a
ti
on
,
a
s foll
ows
2
0
()
2
ee
x
e
S
E
R
F
x
d
x
x
(
22
)
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
User
sel
ect
io
n pro
t
oc
ol in DF
coop
e
ra
ti
ve
ne
t
works wi
th
hy
br
id
T
SR
-
P
SR
p
r
oto
c
ol
ba
se
d… (Ph
u Tra
n Ti
n
)
539
Substi
tuti
ng (1
9) into
(22) a
nd
rep
la
ce
th
x
, we
hav
e:
1
11
0
0
1
2
0
1
2
(
1
)
1
0
11
0
0
1
2
0
1
2
1
0
1
2
0
(
1
)
1
2
(
1
)
2
(
1
)
2
(
1
)
2
(
1
)
2
(
1
)
1
(
1
)
(
2
M
pp
M
x
p
M
pp
x
M
p
x
pp
M
x
x
p
C
M
K
p
e
SER
dx
x
x
x
p
C
M
e
K
p
e
dx
C
M
x
1
1
0
0
1
2
0
(
1
)
1
11
0
0
1
2
0
1
2
0
(
1
)
2
1)
1
(
1
)
(
1
)
2
(
1
)
M
x
p
M
p
p
x
x
M
p
px
e
K
dx
p
px
C
M
e
e
K
dx
p
(
23
)
Her
e
w
e
d
e
note
that
1
0
2
x
e
J
d
x
x
(
24
)
Applyi
ng table
of inte
gr
al
e
q(3.361,
2) in
[18]
We
hav
e
1
2
J
and
we den
ote J
2
a
s
foll
ow
i
ng
1
2
1
1
0
0
1
2
0
1
2
0
1
(
1
)
(
1
)
2
(
1
)
M
p
p
x
M
p
px
J
C
M
e
K
d
x
p
(
25
)
Applyi
ng table
of inte
gr
al
e
q(6.614,
5) in
[18]
, w
e
ha
ve
0
1
2
1
1
2
2
1
0
0
0
1
2
0
1
2
0
1
2
11
(
1
)
4
2
2
p
M
pp
M
p
M
p
p
J
C
e
K
K
(
26
)
(
1
)
1
3
1
1
0
0
1
2
0
1
2
0
1
(
1
)
(
1
)
2
(
1
)
M
p
p
x
x
M
p
px
J
C
M
e
e
K
d
x
p
(
27
)
Finall
y, SER
of the
syst
em
mo
del ca
n
be c
al
culat
ed
as
1
2
3
S
E
R
J
J
J
(
28
)
4.
NUMER
IC
A
L RES
ULTS
AND DIS
C
USSION
In
t
his
sect
ion,
the
Mo
nte
Ca
rlo
sim
ulati
on
is
us
e
d
f
or
validat
ing
the
a
naly
ti
cal
expressio
n
i
n
the
a
bove
sect
ion
.
We
co
ns
ide
r
a
netw
ork
with
on
e
s
our
ce,
one
relay
,
a
nd
m
ulti
-
destin
at
ion
,
wh
e
re
source
-
relay
a
nd
relay
-
destinat
io
n dist
ances a
re
bo
t
h norm
al
iz
ed
to u
nit
value. M
oreo
ver,
we
in
ve
sti
gate an
d
c
om
par
e TSR, P
SR an
d
TPSR cases
in t
he
co
nnect
io
n of al
l p
os
sible
syst
e
m
p
aram
e
te
rs.
The
ef
fect
of
η
on
the
ou
ta
ge
pro
ba
bili
ty
and
syst
e
m
throu
ghpu
t
of
t
he
pro
pose
d
relay
net
work
syst
em
in
the
com
par
ison
betwee
n
T
SR,
PSR,
TP
S
R
cases
are
show
n
in
Fi
gure
3
an
d
4.
I
n
the
se
Fig
ure
s,
we
set
the
m
ai
n
syst
e
m
p
aram
et
ers
as
P
s
/N
0
=
10
dB,
R=
0.5
bps,
a
nd
M=
2.
From
the
res
ults,
the
analy
ti
cal
and
th
e
si
m
ulati
on
res
ults
m
at
ch
fo
r
al
l
po
ssible
va
lues
P
s
/N
0
.
F
ig
ure
3
s
hows
that
the
ou
ta
ge
pro
ba
bili
ty
has
a
consi
der
a
ble
de
crease
w
hile
η
inc
reases
f
ro
m
0
to
1.
O
n
an
ot
her
hand,
syst
e
m
thro
ug
hput
s
ign
ific
a
ntl
y
inc
reases
in
c
onnecti
on
with
i
ncr
easi
ng
of
η
f
ro
m
0
to
1.
T
he
resea
r
ch
res
ults
s
how
t
hat
the
ou
ta
ge
pro
ba
bili
ty
and
th
e
syst
e
m
thr
ough
pu
t
ha
ve
the
be
tt
er v
al
ue
in
t
he
TPS
R ca
se in
co
m
par
iso
n w
it
h
the
oth
e
rs
c
ases.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
2
,
Fe
bru
ary
201
9
:
5
34
–
542
540
Figure
3. O
uta
ge pr
obabili
ty
v
ers
u
s η
Figure
4.
Th
r
ough
pu
t
ve
rsus
η
Figure
5. O
uta
ge pr
obabili
ty
v
ers
us
M
Figure
6.
Th
r
ough
pu
t
ve
rsus
M
Figure
7. O
uta
ge pr
obabili
ty
v
ers
us
P
s
/N
0
Figure
8.
Th
r
ough
pu
t
ve
rsus
P
s
/N
0
Figure
9. Com
par
is
on SER
ve
rsu
s
M
F
igure
10. C
om
par
ison
S
ER
ver
s
us
P
s
/N
0
Evaluation Warning : The document was created with Spire.PDF for Python.
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci
IS
S
N:
25
02
-
4752
User
sel
ect
io
n pro
t
oc
ol in DF
coop
e
ra
ti
ve
ne
t
works wi
th
hy
br
id
T
SR
-
P
SR
p
r
oto
c
ol
ba
se
d… (Ph
u Tra
n Ti
n
)
541
Finall
y,
Fi
g
ure
s
9
a
nd
10
pl
ot
the
e
ff
ect
of
M
and
P
s
/N
0
on
S
ER
in
the
c
om
par
iso
n
bet
ween
TSR,
PSR,
TSPR
ca
ses
wi
th
the
m
ai
n
pa
ram
et
ers
of
P
s
/N
0
=1
0,
η=
0.8
,
and
M=
2.
The
SER
decr
ease
s
w
hile
M
an
d
P
s
/N
0
increases
f
ro
m
1
t
o
10
a
nd 0
t
o
20, r
es
pecti
ve
ly
.
The
res
ults
sho
w
t
hat
al
l sim
ulati
on
a
nd
analy
ti
cal
res
ul
ts
are
m
at
ched
w
el
l
with eac
h
oth
e
r
.
M
or
e
over
,
S
ER i
n
case
TS
PR is bett
er
tha
n
the
rem
ai
nin
g
cases
.
Fu
rt
her
m
or
e,
the
ou
ta
ge
pro
bab
il
it
y
an
d
s
yst
e
m
thro
ug
hput
of
the
pr
opose
d
syst
em
ver
s
us
M
are
il
lustrate
d
in
th
e
Fig
ure
5
a
nd
6
with
P
s
/N
0
=
10
dB
an
d
R=
0.5
bps.
From
the
res
ults,
we
sh
ow
that
th
e
ou
ta
ge
pro
bab
il
it
y
inc
reases
a
nd
t
he
syst
e
m
throughput
decr
ea
se
r
e
m
ark
ably
with
i
ncr
easi
ng
M
from
0
t
o
10.
On
the
sam
e
way,
t
he
ou
ta
ge
prob
a
bi
li
ty
and
the
sys
tem
thro
ug
hput
ver
s
us
P
s
/N
0
with
R=
0.5
bp
s
an
d
M=
2
as
s
how
n
in
Fig
ure
s. 7 a
nd 8.
All t
he
a
naly
ti
cal
an
d
si
m
ula
ti
on
r
e
su
lt
s
ag
ree
well
w
i
th each
ot
her
.
5.
CONCL
US
I
O
N
In
this
w
ork
,
w
e
in
vestigat
e
t
he
syst
em
per
f
orm
ance
of
the
e
nergy
harvesti
ng
f
ull
-
duple
x
(
FD
)
dec
od
e
-
and
-
f
orwa
rd
(
DF
)
hybri
d T
S
R
-
PSR
pr
oto
c
ol
relay
ing
net
w
ork.
In
t
he se
le
ct
ion
sc
hem
e,
t
he best
us
er
s
el
ect
ion
protoc
ol
is
pro
po
s
ed
a
nd
in
ve
sti
gated.
Ma
inl
y
we
de
rive
the
cl
os
ed
-
f
orm
e
xpressi
on
f
or
t
he
ou
ta
ge
probabil
it
y,
syst
e
m
throu
ghput
a
nd
t
he
sym
bo
l
error
rate
(SER
)
of
t
he
s
yst
e
m
.
Nu
m
erical
res
ults
are
al
so
pr
e
sente
d
by
the
Mon
te
r
Ca
rlo
s
i
m
ulati
on
to
va
li
date
the
the
oret
ic
al
analy
sis
in
c
onnecti
on
with
t
he
al
l
possible
par
am
eter
s
i
n
the
com
par
iso
n
bet
wee
n
TP
S
R,
TS
R
an
d
P
SR
cases.
T
he
researc
h
resu
lt
s
show
that
TP
SR
case
is
bett
er
tha
n
the
oth
e
rs
in
t
erm
of
outa
ge
pr
ob
a
bili
ty
an
d
SER.
T
his
pa
per
pro
vid
es
the
novel
rec
om
m
end
at
ion
f
or
th
e
com
m
un
ic
at
ion
relay
ing
net
work
in
the
ne
ar futu
re
.
REFERE
NCE
S
[1]
Chen,
He,
Cha
o
Z
h
ai
,
Yongh
ui
L
i,
and
Br
anka
Vuce
t
ic.
"Coopera
ti
v
e
S
tra
t
egi
es
for
W
ire
le
ss
-
Pow
ere
d
Com
m
unic
at
ions:
An
Overvi
ew.
"
IE
EE
Wirel
ess
Com
municat
ions
2
5,
no
.
4
(0
8
2018):
112
-
19.
doi:
10.
1109
/mw
c.
2017.
1700245
.
[2]
Bi,
S.,
Ho,
C.
K.,
&
Zha
ng
,
R.
W
ireless
powere
d
communic
a
ti
on:
Opp
ortuni
ties
and
cha
l
le
nges.
I
EEE
Comm
unic
ati
ons Magazine
,
2015
;
53
:
117
-
125.
[3]
Ni
y
a
to,
D.,
K
im,
D.
I.,
Maso,
M.,
&
Han,
Z
.
W
ireless
Pow
ere
d
Co
m
m
unic
at
ion
Ne
tworks:
Resea
r
c
h
Dire
c
ti
ons
and
Te
chno
logi
c
al Approac
hes.
IE
EE Wi
rel
ess Comm
unic
ati
ons
,
2017
,
2
-
11
.
[4]
Yu,
H.,
Le
e
,
H.
,
&
Jeon,
H.
(20
17,
10).
W
ha
t
is
5G
?
Emerging
5G
Mobile
Serv
ic
es
and
Networ
k
Requi
rement
s.
Sus
ta
ina
bilit
y
,
2
017,
9
,
1848
.
[5]
C.
R.
Va
le
nt
a
an
d
G.
D.
Durgin,
‘‘Harve
sting
wir
el
ess
power:
Sur
ve
y
of
en
erg
y
-
h
arv
este
r
conve
rs
ion
eff
i
ci
e
nc
y
i
n
far
-
field,
wire
le
s
s power
tr
ansfe
r s
y
stems
,
’’
I
EE
E
Mic
row.
Mag
.
,
2
014
;
15
(
4
):
108
-
120
.
[6]
La
i
,
Xi
az
h
i,
W
a
nxin
Zou
,
Xutao
Li,
and
Li
sh
eng
Fan.
"M
ult
iuser
En
erg
y
Harv
est
ing
Re
lay
i
ng
S
ystem
with
Di
rect
Li
nks."
I
ET
Co
mm
unic
ati
ons
1
1,
no
.
12
(08, 20
17):
1846
-
852.
d
oi:
10.
1049
/i
e
t
-
c
om
.
2016.
1288.
[7]
Kawaba
ta,
Hirok
i,
and
Koji
Ishib
ashi.
"
R
F
Ene
rg
y
Powe
red
F
ee
d
back
-
aid
ed
Coo
perati
on.
"
2014
IEE
E
25th
Annu
al
Inte
rna
ti
ona
l
Sy
m
posium
on
Personal,
Indoo
r,
and
Mobile
Rad
io
Com
muni
cation
(PIM
RC)
,
09
201
4
.
doi:
10.
1109
/pi
m
rc.
2014
.
713
6211
.
[8]
Krikidi
s,
Io
anni
s
,
Ste
li
os
T
imotheou,
and
Shig
en
obu
Sasaki.
"RF
Ene
rg
y
Tr
ansfe
r
for
Cooper
at
i
v
e
Networks:
Da
t
a
Rel
a
y
ing
or
Ene
rg
y
Harve
s
ti
ng
?
"
IE
EE
Comm
unic
ati
ons
Lett
ers
16,
no.
11
(11
2
012):
1772
-
77
5.
doi:
10.
1109
/l
co
m
m
.
2012.
091712.
121395.
[9]
Nasir
,
Al
i
A.
,
Xi
ang
y
un
Zhou
,
Salman
Durran
i,
a
nd
Rodne
y
A
.
K
enne
d
y
.
"Rel
a
y
i
ng
Protocol
s
for
W
ire
le
ss
En
er
g
y
Harve
sting
and
I
nform
at
ion
Proc
essing.
"
IEEE
Tr
ansacti
ons
on
Wi
rele
ss
Comm
unic
ati
ons
12,
no.
7
(07
2013)
:
3622
-
636.
doi
:10.
1109
/t
wc.
2013
.
06241
3.
122042.
[10]
La
neman,
J
.
n.
,
D.n.
c
.
Tse
,
and
G.w.
W
orne
ll
.
"
Coopera
t
ive
Div
ersity
in
W
ireles
s
Networks:
Eff
i
ci
en
t
Proto
col
s
a
nd
Outage
B
eha
v
i
or.
"
IE
EE
Tr
ansacti
ons
on
Information
Theory
50,
n
o.
12
(
12
2
004
):
3062
-
08
0.
doi:
10.
1109
/t
it.2
004.
838089.
[11]
Micha
lopoul
os
,
Diom
idi
s
S.,
H
imal
A
.
Surawee
ra
,
and
Rob
e
rt
Schob
er.
"Re
lay
S
el
e
ction
f
or
Sim
ult
an
eou
s
Inform
at
ion
Tr
a
nsm
ission
and
W
ire
le
ss
Ene
rg
y
Tra
nsfer:
A
Tradeoff
Perspec
ti
v
e.
"
IE
EE
Journal
on
Selec
te
d
Ar
eas
in
Comm
unic
at
i
ons
,
2015,
1.
doi
:10.
1109/j
s
ac
.
20
15.
2391771.
[12]
Ta
n
N.
Ngu
y
en,
T.
H.Q.Minh
,
Phuong
T
.
Tra
n
and
Miroslav
Vozn
a
k.
Ene
rg
y
Harv
e
sting
ov
er
R
ic
i
an
Fading
Chann
el
:
A
Perform
anc
e
Anal
y
sis
for
Ha
lf
-
Duplex
Bidi
r
e
ct
ion
al
Sensor
Networks
under
Hardwar
e
Im
pai
rm
ent
s.
S
ensors
,
2018,
18
.
[13]
Ta
n
N.
Ngu
y
en
,
T.
H.Q.Minh
,
Ph
uong
T.
Tr
an
an
d
Miroslav
Vozn
ak.
Adap
ti
v
e
En
erg
y
Harv
esti
ng
Rel
a
y
ing
Protoc
ol
for
Two
-
W
a
y
H
al
f
Duplex
S
y
s
t
em
Network
ov
er
R
ic
i
an
Fadin
g
Chann
el.
Wire
le
ss
Comm
unic
a
ti
ons
and
Mob
ile
Computing
,
201
8.
[14]
Ti
n,
Phu
Tra
n
,
T
ran
Hoang
Quan
g
Minh,
Ta
n
N
.
Ngu
y
en
,
and
Mi
ros
la
v
Vozna
k
.
"
S
y
stem
Perform
anc
e
Anal
y
sis
of
Half
-
Duplex
Re
lay
N
et
work
o
ver
Ri
cian
Fad
ing
Chann
el."
TEL
KOMNIKA
Tele
communic
a
t
ion,
Computing
,
El
e
ct
ronics
and
Control
,
16
,
no
.
1
(
02,
2018
):
18
9.
doi
:10.
12928
/
te
lkomnika
.
v16i
1.
7491.
[15]
Rashid,
T
ari
qu
e,
Sunil
Kum
ar
,
Aks
ha
y
Verm
a,
Prate
ek
R
aj
G
au
ta
m
,
and
Arv
ind
Kum
ar.
"P
m
-
EE
MRP
:
Pos
tura
l
Movem
ent
Base
d
En
erg
y
Eff
i
cient
Multi
-
hop
R
outi
ng
Proto
col
for
Intr
a
W
ire
l
e
ss
Bod
y
Sensor
Network
(Intr
a
-
W
BS
N)."
TEL
K
OMNIKA
Telec
omm
unic
ati
on,
Computing,
Ele
ct
ronics
and
Co
ntrol
16
,
no.
1
(02,
2018
):
166
.
doi:
10.
12928
/t
e
l
kom
nika
.
v16i1.
7
318.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2502
-
4752
Ind
on
esi
a
n
J
E
le
c Eng &
Co
m
p
Sci,
Vo
l.
1
3
, N
o.
2
,
Fe
bru
ary
201
9
:
5
34
–
542
542
[16]
A.
F.
Morabit
o
,
"P
ower
S
y
nthe
si
s of
Mask
-
Constrai
ned
Shaped
Bea
m
s T
hrough M
axi
m
al
l
y
-
Spar
se
Plana
r
Arra
y
s
,
"
Telk
omnika
(
Tele
communic
at
ion
Computing Elec
tronic
s and
Con
t
rol)
,
2016
;
14
(
4
)
:
1217
-
1219
.
[17]
Ngu
y
en
,
T
.
N.
,
T.
T.
Du
y
,
G.
-
T.
Luu,
P.
T.
Tra
n
,
and
M.
Vozna
k
.
"Ene
rg
y
H
arv
es
ti
ng
-
base
d
Spectrum
Acc
ess
with
Inc
rement
a
l
Coo
per
ation,
R
el
a
y
Sele
c
ti
on
and
Hardware
Noises.
"
Radi
oengi
n
ee
r
ing
26,
no
.
1
(0
4,
2017):
240
-
5
0
.
doi:
10.
13164
/re
.
2017.
0240.
[18]
Ta
bl
e
o
f
Inte
gr
als
,
Serie
s
,
an
d
Pr
oduct
s,
2015
.
[19]
Mcka
y
,
Matt
h
e
w
R.
,
Alex
J.
Grant
,
and
Iai
n
B.
Coll
ings
.
"
Perform
anc
e
An
aly
s
is
of
MIM
O
-
MRC
in
Double
-
Corre
lated
Ra
y
l
ei
gh
Envi
ronm
e
nts."
IE
EE
Tr
ansacti
ons
on
C
omm
unic
ati
ons
55,
no
.
3
(
03
2007
):
497
-
507
.
doi:
10.
1109
/t
co
m
m
.
2007.
892450.
[20]
Tra
n
Ho
ang
Qu
ang
Minh.
"H
y
b
rid
T
ime
-
Pow
er
Sw
it
chi
ng
Proto
col
of
En
erg
y
H
arv
esti
ng
Bidi
r
e
ct
ion
al
R
ela
y
ing
Network:
Throu
ghput
and
Ergod
ic
Cap
acit
y
Ana
l
y
sis."
TEL
KOMNIKA
Telecomm
unic
ati
on
,
Comp
uti
ng,
Elec
troni
c
s
and
Control
,
16,
no.
5
(10
,
2018)
:
189.
htt
p
:/
/dx
.
d
oi.
org/10
.
12928/
te
lkomnika
.
v16i
5
.
9118.
Evaluation Warning : The document was created with Spire.PDF for Python.