Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
9
, No
.
5
,
Octo
ber
201
9
, pp.
3399
~3
406
IS
S
N: 20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v9
i
5
.
pp33
99
-
34
06
3399
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
H
alf
-
du
plex p
ower be
acon
-
assisted
energy
harvesti
ng rel
ay
i
ng
network
s: syste
m perfo
rm
ance a
nalysis
Tan N.
N
guye
n
1
,
Minh
Tra
n
2
, Van
-
D
uc Ph
an
3
,
H
oang
-
N
am Nguye
n
4
,
Thanh
-
Lo
ng
Nguy
e
n
5
1
W
ire
le
ss
Com
muni
cations
Rese
a
rch
Group,
Facult
y
of El
ec
tr
ic
a
l a
nd
Elec
tron
ic
s
E
ngine
er
ing,
Ton
Duc Than
g
Univer
sit
y
,
Vie
t
nam
2
Optoel
e
ct
roni
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
Cent
er
of E
xcel
le
nc
e
for
Autom
at
ion
and
Prec
isi
on
Mec
han
ical E
ngine
er
ing, Nguy
en
T
at T
hanh
U
nive
rsit
y
,
Vi
et
na
m
4
Modeli
ng
Evolutiona
r
y
A
lgori
t
hm
s Sim
ula
ti
on
and
Artif
ic
i
al Int
el
li
g
ence,
Facul
t
y
of Electr
ic
a
l
&
E
le
c
tronics E
ngin
ee
rin
g,
Ton
Duc Thang
Univer
sit
y
,
Vie
t
na
m
5
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
N
ov
11
, 201
8
Re
vised
A
pr
6
,
20
9
Accepte
d
Apr
18
, 201
9
In
thi
s
work,
t
he
hal
f
-
dup
le
x
(HF
)
power
bea
con
-
assisted
(P
B)
ene
rg
y
har
vesti
ng
(EH)
rel
a
y
i
ng
net
wor
k,
which
consist
s
of
a
source
(S),
Rela
y
(R)
,
desti
nation
(D)
and
a
pow
er
be
ac
on
(PB)
a
re
i
ntroduc
ed
and
i
nvesti
gated.
Firstl
y
,
th
e
an
aly
t
ical
expr
essions
o
f
the
s
y
st
e
m
per
form
anc
e
in
te
rm
of
outa
ge
prob
abi
l
i
t
y
(OP
)
and
th
e
s
y
stem
throughput
(ST)
are
an
aly
z
ed
and
der
ive
d
in
both
amplif
y
-
and
-
fo
rward
(AF
)
and
dec
od
e
-
and
-
for
ward
(DF
)
m
odes.
After
tha
t,
we
ver
if
y
th
e
cor
re
ct
ness
of
the
ana
l
y
t
i
ca
l
ana
l
y
sis
b
y
usin
g
Monte
-
C
arl
o
sim
ula
t
ion
in
conn
ec
t
ion
with
th
e
pri
m
ar
y
s
y
s
te
m
par
amete
rs
.
Fro
m
the
num
erica
l
result
s,
we
ca
n
see
th
at
all
th
e
a
naly
tica
l
and
the
sim
ulation re
sults a
re
m
at
ch
e
d
well wit
h
each
othe
r
.
Ke
yw
or
d
s
:
Half
-
duple
x (HD)
Ou
ta
ge pr
obab
il
ity (O
P
)
Syst
e
m
thr
oughput
(S
T
)
W
i
reless e
nerg
y har
vestin
g
(EH)
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
s
:
Van
-
D
uc
P
han
,
Ce
nter of
Ex
c
e
ll
ence for
A
utom
at
ion
and Pre
ci
sion
Mec
han
i
cal
Enginee
rin
g,
Nguyen
Tat T
ha
nh Unive
r
sit
y,
Ho Chi Mi
nh
Ci
ty
, V
ie
tnam
.
Em
a
il
:
pv
duc
@n
tt
.e
du.vn
Ho
a
ng
-
Nam
N
gu
ye
n
,
Mod
el
in
g
E
voluti
on
a
ry
Algor
it
h
m
s
Si
m
ulatio
n an
d A
rtific
ia
l In
te
ll
igence
,
Faculty
of Elec
tric
al
&
Ele
ct
r
on
ic
s
Engine
er
ing
,
To
n Du
c
Th
a
ng
Un
i
ver
sit
y,
Ho Chi Mi
nh
Ci
ty
, V
ie
tnam
.
Em
a
il
:
ng
uyen
ho
a
ngnam
@td
tu.edu.
vn
1.
INTROD
U
CTION
Energy
ha
rv
e
s
ti
ng
(E
H)
relay
network,
w
hi
ch
us
es
a
ra
di
o
fr
e
quency
(
RF)
sig
nal
for
wirelessl
y
trans
fer
r
in
g
po
wer,
has
at
trac
te
d
m
uch
at
te
nt
ion
beca
us
e
of
prolo
ngin
g
the
li
fetim
e
of
a
wireless
net
work.
This
s
olu
ti
on
c
an
be
ob
ta
in
ed
beca
us
e
RF
si
gn
al
ca
n
sim
ul
ta
neously
tran
s
fer
e
ne
rg
y
a
nd
inf
or
m
at
ion
[
1
-
10]
.
Nowa
days,
th
e
syst
e
m
per
f
or
m
ance
of
EH
relay
in
g
ne
twork
has
be
en
stu
died
i
n
m
any
stud
ie
s
.
[
11
]
inv
est
igate
d
th
e
fu
ll
-
du
plex
EH,
the
dev
el
opm
ent
of
coop
erati
ve
prot
ocol
s
fo
r
E
H
rela
yi
ng
netw
ork
i
s
fu
ll
y
stud
ie
d
in
[
12
-
13
]
.
F
urt
her
m
or
e
,
[
14
-
15]
in
tro
du
ce
a
nd
in
vestigat
e
a
“ha
rv
est
-
the
n
-
tra
nsm
it
”
pr
oto
c
ol
for
a
m
ul
ti
-
us
er
rela
yi
ng
net
wor
k.
In
al
l
pa
pers
a
bove,
t
he
rela
y
(R)
an
d
the
destinat
io
n
(
D
)
node
s
only
r
ecei
ve
energy
f
ro
m
the
s
ource
(S)
o
r
the
acce
ss
point
nodes.
In
t
he
tre
nds
to
im
pr
ov
i
ng
EH
a
nd
in
for
m
at
ion
transm
issi
on
(IT)
proces
ses
in
the
wireless
relay
network,
so
m
e
research
ers
pro
posed
t
he
idea
of
de
pl
oying
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
5
,
Oct
ober
20
19
:
3
3
9
9
-
3
4
0
6
3400
ded
ic
at
e
d
pow
er
beac
on
no
de
(P
B).
I
n
the
wireless
relay
network
with
us
ing
th
e
PB
node,
D
an
d
R
can
receive
e
ne
rg
y
both
f
ro
m
S
a
nd
PB
node
s
[
16
-
17
]
.
The
pr
ob
le
m
of
ene
r
gy
ha
r
vestin
g
i
s
al
so
enc
ount
ered
in
directi
onal
MA
NET
c
onte
xts i
nvolv
i
ng h
i
gh
directi
ve beam
form
ing
d
e
vice
s [18
]
.
In
t
his
w
ork
,
we
intr
oduc
e
a
nd
in
vestigat
e
the
sys
te
m
per
form
ance
analy
sis
of
H
D
po
wer
beac
on
-
assist
ed
EH
re
la
y
netwo
r
k
in
bo
th
t
he
am
pl
ifie
r
-
a
nd
-
f
orwa
rd
(
AF)
an
d
de
cod
e
-
a
nd
-
f
orward
(
DF)
m
o
des
vi
a
the
Ra
le
igh
fa
ding
cha
nn
el
s
.
Firstl
y,
the
integral
cl
os
e
d
-
f
orm
exp
ressio
n
of
the
ou
ta
ge
pro
bab
il
it
y
(OP)
an
d
syst
e
m
throughput
(S
T
)
.
A
fter
that,
the
analy
ti
cal
ex
pr
essi
ons
a
re
co
nv
i
nced
by
us
in
g
M
on
t
e
-
Ca
rlo
si
m
ulati
on
with
hel
ping
Ma
t
Lab
s
of
twa
re
in
both
am
plifie
r
-
a
nd
-
f
orwa
rd
(A
F
)
an
d
dec
ode
-
a
nd
-
forw
a
r
d
(
DF)
m
od
es.
Finall
y,
the
num
eri
cal
analy
sis
can
be
dem
onstr
at
ed
in
c
onnecti
on
with
the
pri
m
ary
s
yst
e
m
par
am
et
er
.
2.
SY
STE
M MO
DEL A
N
D
PE
RFO
R
MANCE
A
N
AL
YS
IS
Fig
ure
1
il
lustrate
s
the
syst
em
m
od
el
of
t
he
pr
opos
e
d
s
yst
e
m
.
In
Fig
ur
e
1,
t
he
inf
or
m
at
ion
is
trans
ferred
fro
m
S
to
D
with
helpi
ng
of
an
interm
ediat
e
R.
Here
S,
a
nd
R
are
ha
rv
e
s
t
energy
f
ro
m
the
PB
node
dir
ect
ly
.
In
this
m
od
e
l,
al
l
the
blo
c
k
-
fad
i
ng
c
hannels
are
the
Ra
le
igh
fad
i
ng
cha
nn
el
s
.
Fig
ure
2
il
lustrate
s
the
energy
harvest
ing
a
nd
i
nfor
m
at
ion
tra
ns
m
issi
on
proces
ses.
In
this
pro
po
se
d
syst
em
m
od
el
,
the
node
PB
tran
sf
ers
the
e
ne
rg
y
to
S
a
nd
R
in
α
T
(
≤
α
≤
1)
.
Af
te
r
that,
S
tr
ansf
e
rs
t
he
in
f
or
m
at
ion
to
R
in
the
nex
t
i
nterval
ti
m
e
(1
-
α
)
T/2.
F
inall
y,
the
rela
y
node
R
tra
nsfers
t
he
in
form
at
ion
to
t
he
de
sti
nation
node
D
in
(1
-
α
)T/2
.
Figure
1. Syst
em
m
od
el
Figure
2. EH
a
nd
IT
pr
ocesse
s
2.1.
The am
pli
fier
an
d
forw
ard (
AF) m
od
e
In
α
T,
t
he
PB
node
tra
ns
fe
rs
energy
to
both
so
urce
no
de
S
an
d
relay
no
de
R.
Th
en
th
e
harveste
d
energy at
the
s
ource
node
S
c
an be cal
c
ulate
d by
2
s
B
B
S
E
T
P
h
(1)
w
he
re
0<
η
<
1
is
the
e
nergy
c
onve
rsion
e
ff
ic
ie
ncy.
F
ro
m
th
is
ene
rg
y,
the
a
ver
a
ge
t
ran
sm
it
s
powe
r
at
S
c
an
be
form
ulate
d
as
2
(
1
)
/
2
s
s
B
B
S
E
P
k
P
h
T
(2)
Si
m
il
arity, the har
vested
en
e
r
gy at R i
s
2
r
B
B
R
E
TP
h
(
3)
The
n
the
av
e
ra
ge
tra
ns
m
it
s p
ower
at
R is
2
(
1
)
/
2
r
r
B
B
R
E
P
k
P
h
T
(
4)
w
he
re
we de
note
that
2
1
k
.
S
ou
r
c
e
(
S
)
Re
l
ay
(
R
)
De
s
t
in
at
ion
(
D
)
P
owe
r
B
e
ac
on
(
PB
)
BR
h
SR
h
RD
h
BS
h
E
H
a
t
S
,
R
I
n
f
or
m
at
i
on
T
r
an
s
m
i
s
s
i
on
R
t
o
D
α
T
(
1
-
α
)
T
/
2
(
1
-
α
)
T
/
2
T
I
n
f
or
m
at
i
on
T
r
an
s
m
i
s
s
i
on
S
t
o
R
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Ha
lf
-
duplex
po
we
r
be
acon
-
assi
ste
d
ene
rg
y
harvesti
ng rel
ay
ing
netwo
rk
s: s
yst
em
.
..
(
T
an
N. Ng
uyen
)
3401
Af
te
r
that,
S tr
ansf
e
rs
t
he
in
f
or
m
at
ion
to
R
. T
he recei
ve
d
si
gn
al
is
r
S
R
s
r
y
h
x
n
(5)
wh
e
re
x
s
is t
he t
ran
sm
issi
on
si
gn
al
from
S.
In
the
nex
t
st
age,
R
trans
fe
rs
the
inf
orm
a
ti
on
,
w
hic
h
is
received
fro
m
the
so
ur
ce
to
D
in
the
rem
ai
nin
g
inter
val tim
e. Th
e
r
ecei
ved
sig
nal
at
D
is
form
ulate
d
as
d
R
D
r
d
y
h
x
n
(6)
w
he
re
n
r
,n
d
a
r
e
the
ad
diti
ve
wh
it
e
Ga
us
si
an
noise
(
A
W
GN)
at
R,
D
with
zer
o
m
ea
n
an
d
var
ia
nc
e
N
0
,
2
sr
xP
,
2
rr
xP
, and
is ex
pe
ct
at
ion
ope
rato
r,
x
r
is t
he
tra
nsm
issi
on
sig
nal
from
R.
In the
AF m
od
e, w
e
can
calc
ul
at
e the am
plifi
er f
act
or as
2
0
rr
r
s
S
R
xP
y
P
h
N
(7)
Fr
om
(
6) a
nd (7), the
r
ecei
ve
d
si
gn
al
at
D
is
d
R
D
r
d
R
D
S
R
s
r
d
S
R
R
D
s
R
D
r
d
s
ig
n
a
l
n
o
is
e
y
h
y
n
h
h
x
n
n
h
h
x
h
n
n
(8)
Using
(8),
we
can calc
ulate
the si
gn
al
t
o no
i
se r
at
io
(S
R
N)
at
D
by
2
2
2
2
2
2
22
2
2
0
0
0
0
0
()
S
R
R
D
s
S
R
R
D
s
ee
R
D
s
S
R
RD
r
h
h
P
h
h
P
h
N
N
N
P
h
N
hN
P
(10)
The
e
quat
ion (
10)
is
refo
rm
ul
at
ed
by
us
i
ng
N
0
<<P
r
22
2
22
00
S
R
R
D
s
r
ee
R
D
r
s
S
R
h
h
P
P
h
P
N
N
P
h
(11)
Com
bin
ing
with
(2),(4
)
a
nd s
ub
sti
tuti
ng i
nto (
11),
SN
R i
s
2
2
2
2
0
2
2
2
2
2
00
B
S
R
R
D
B
R
B
S
ee
R
D
B
R
B
S
S
R
k
P
h
h
h
h
k
X
Y
XY
h
h
N
N
h
h
(
12)
w
he
re
we de
note
0
0
B
P
N
,
2
2
2
2
,
R
D
B
R
S
R
B
S
X
h
h
Y
h
h
.
Fo
r
m
or
e analy
sis, u
ti
li
zi
ng th
e res
ult i
n [19],
the
c.
d.f.
of
X and Y
is
1
(
)
1
2
2
X
R
D
B
R
R
D
B
R
F
x
x
K
x
(13)
1
(
)
1
2
2
Y
S
R
B
S
S
R
B
S
F
y
y
K
y
(14)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
5
,
Oct
ober
20
19
:
3
3
9
9
-
3
4
0
6
3402
OP
of the
prop
os
e
d
syst
em
is
0
20
00
0
P
r
(
)
P
r
P
r
,
Pr
1,
AF
o
u
t
e
e
th
th
th
th
th
th
th
th
k
X
Y
P
X
k
Y
Y
XY
Y
XY
k
Y
k
Y
k
(15)
w
he
re
21
R
th
is a th
re
sh
ol
d, an
d
R i
s
sour
ce
rate.
Propos
it
io
n
1
:
OP
of the
prop
os
e
d
syst
em
is
giv
e
n by
0
01
00
1
4
(
2
)
2
th
AF
R
D
B
R
th
R
D
B
R
th
o
u
t
S
R
B
S
S
R
B
S
th
th
k
yy
P
K
y
K
dy
k
y
k
y
(16)
Proof
: See
Appendix
A.
Finall
y, The
syst
e
m
thr
ough
put at
D
is f
orm
ulate
d
as
0
0
0
1
0
(
2
)
(
1
)
(
1
)
(
1
)
(
1
)
2
(
1
)
22
2
th
R
D
B
R
th
S
R
B
S
th
A
F
A
F
A
F
o
u
t
o
u
t
S
R
B
S
R
D
B
R
th
k
th
y
Ky
ky
R
T
R
P
P
R
T
y
K
dy
ky
(17)
2.2.
The
dec
od
e
and f
orwar
d
(
DF) m
od
e
Fr
om
(
1) a
nd (2) we
h
a
ve
t
he SNR
for
t
he
D
F m
od
e:
2
2
2
10
00
s
S
R
B
B
S
S
R
P
h
k
P
h
h
S
N
R
k
Y
NN
(18)
2
2
2
20
00
r
R
D
B
B
R
R
D
P
h
k
P
h
h
S
N
R
k
X
NN
(19)
Fu
rt
her
m
or
e,
OP
of the
prop
os
e
d
syst
em
is
1
2
0
0
Pr
m
in
,
Pr
m
in
,
DF
o
u
t
th
th
P
S
N
R
S
N
R
k
Y
k
X
(20)
Propos
it
io
n
2
:
The o
utage pr
obabili
ty
at the
destinat
io
n nod
e of the
prop
ose
d
syst
em
is g
iven
b
y:
11
0
0
0
1
4
2
2
DF
t
h
R
D
B
R
t
h
S
R
B
S
t
h
o
u
t
R
D
B
R
S
R
B
S
P
K
K
k
k
k
(21)
Proof
: See
Appendix
B.
Finall
y, the t
hroug
hput
τ
at
D
is give
n by:
11
0
0
0
(
1
)
(
1
)
2
(
1
)
2
2
2
D
F
D
F
t
h
R
D
B
R
t
h
S
R
B
S
t
h
o
u
t
R
D
B
R
S
R
B
S
R
P
R
K
K
k
k
k
(22)
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Ha
lf
-
duplex
po
we
r
be
acon
-
assi
ste
d
ene
rg
y
harvesti
ng rel
ay
ing
netwo
rk
s: s
yst
em
.
..
(
T
an
N. Ng
uyen
)
3403
3.
RESU
LT
S
A
ND
DI
SCUS
S
ION
We
us
e
t
he
Mon
te
Ca
rl
o
sim
ulati
on
to
ver
i
f
y
the
correct
ne
ss
of
t
he
analy
ti
cal
exp
ressi
on
of
the
OP
and
ST
in
the
ab
ove
sect
io
n
in
t
he
c
onnec
ti
on
of
the
pri
m
ary
par
am
et
e
rs
of
the
pr
opos
e
d
syst
em
.
Othe
r
si
m
ulati
on
p
a
r
a
m
et
ers
are li
ste
d
in
Ta
ble 1.
Table
1.
Sim
ul
at
ion
par
am
et
e
rs
Sy
m
b
o
l
Na
m
e
Valu
es
η
Energy ha
rves
tin
g
ef
f
icien
cy
0
.7
BS
Mean o
f
2
BS
h
0
.5
BR
Mean o
f
2
BR
h
0
.5
SR
Mean o
f
2
SR
h
0
.5
RD
Mean o
f
2
RD
h
0
.5
th
SNR th
resh
o
ld
7
P
B
/N
0
So
u
rce
p
o
wer
to
n
o
ise
ratio
0
-
20dB
R
So
u
rce
rate
1
.5 b
it/s/Hz
Fig
ure
3
an
d
F
igure
4
pl
ot
the
curves
of
O
P
and
ST
ver
s
us
α
fo
r
both
AF
and
DF
m
od
es.
It
is
sho
wn
in
Fig
ure
3
a
nd
Fig
ure
4
t
hat
P
B
/N
0
at
10
a
nd
20
dB
,
an
d
α
var
ie
s
f
r
om
0
to
1.
We
ca
n
see
from
the
resu
lt
s
that
OP
has
a
decr
ea
se
w
hen
α
increase
f
r
om
0
to
1
as
show
n
Fi
g
ure
3
.
M
or
e
ov
e
r,
Fig
ure
4
s
hows
t
ha
t
the
thr
oughput
inc
reases
in
the
fi
rst
sta
ge
wh
e
n
α
increases
from
0
to
the
optim
a
l
value.
A
fter
that,
the
syst
e
m
thr
oughput
has
a
sig
nifica
nt
de
crease
w
hile
α
var
ie
s
f
ro
m
op
t
im
al
value
t
o
1.
I
n
both
Fig
ure
3
an
d
Fig
ur
e
4,
the a
naly
ti
cal
an
d si
m
ulati
on
re
su
lt
s agree
ve
ry w
el
l
with ea
ch othe
r.
Figure
3.
O
uta
ge pr
obabili
ty
v
ers
us
α
Figure
4. Th
r
ough
pu
t
ve
rsus
α
Fig
ure
5
an
d
F
igure
6
il
lustrat
e the ef
fect
of
η on
OP
a
nd
S
T o
f
t
he
m
od
el
syst
e
m
. H
ere
we
set
α=0.
5
and
P
B
/N
0
at
10
an
d
20
dB,
r
especti
vely
.
F
r
om
the
resu
lt
s,
OP
decr
ease
s
wh
il
e
η
inc
re
ases
from
0
to
1
as
sh
ow
n
in
Fig
ure
5.
I
n
co
ntrast
,
ST
has
a
co
nsi
der
a
ble
i
m
pr
ovem
ent
wh
en
η
increases
f
rom
0
to
1
as
sh
own
i
n
Fig
ure
6.
For
both ca
ses, t
he
a
naly
ti
cal
si
m
ul
at
ion
resu
lt
s a
r
e the sam
e v
al
ues.
M
or
e
over
,
O
P
and
ST
ve
rsu
s
R
are
show
n
in
Fig
ure
7
an
d
Fig
ur
e
8,
r
especti
vely
.
Sim
il
arity,
we
set
P
B
/N
0
at
10
an
d
α
at
0.
2,
0.8
.
Fr
om
Fig
ure
7
we
see
that
OP
inc
reases
wh
il
e
R
var
ie
s
fr
om
1
to
8.
In
c
on
tra
st,
S
T
increase
in
t
he
first
interval
R
to
optim
al
valu
e
then
has
a
huge
decr
ease
as
s
how
n
in
Fig
ure
8.
F
ur
th
erm
or
e
,
Fig
ure
10
sho
ws
the
op
ti
m
al
tim
e
s
witc
hing
fact
or
of
the
pro
po
sed
s
yst
em
a
t
R=
1
an
d
3
bps.
In
al
l
f
ig
ure
s,
the
sim
u
la
tio
n
a
nd
analy
ti
cal
r
esul
ts agr
ee
w
it
h e
ach
oth
e
r.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
5
,
Oct
ober
20
19
:
3
3
9
9
-
3
4
0
6
3404
Figure
5.
O
uta
ge pr
obabili
ty
v
ers
us
η
Figure
6. Th
r
ough
pu
t
ve
rsus
η
Figure
7.
O
uta
ge pr
obabili
ty
v
ers
us
R
Figure
8. Th
r
ough
pu
t
ve
rsus
R
Figure
9. O
uta
ge pr
obabili
ty
v
ers
us
PB/
N
0
Figure
10. T
hr
oughput
ver
s
us PB/
N0
Figure
11. O
ptim
al
tim
e
switc
hing
factor ve
r
su
s
PB/
N0
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Ha
lf
-
duplex
po
we
r
be
acon
-
assi
ste
d
ene
rg
y
harvesti
ng rel
ay
ing
netwo
rk
s: s
yst
em
.
..
(
T
an
N. Ng
uyen
)
3405
4.
CONCL
US
I
O
N
In
t
his
pap
e
r,
we
in
vestigat
e
the
H
D
PB
EH
relay
netw
ork
in
both
A
F
a
nd
DF
m
od
es
.
W
e
der
i
ve
t
he
cl
os
ed
-
f
or
m
ex
pr
essi
ons
of
O
P
an
d
S
T
of
t
he
m
od
el
syst
em
.
Moreo
ve
r,
the
analy
ti
cal
analy
sis
is
co
nvince
d
total
ly
by
the
Mon
te
Ca
rl
o
si
m
ulati
on
.
Also
,
the
op
ti
m
al
tim
e
swit
chin
g
fact
or
i
s
inv
est
igate
d
.
All
the
analy
ti
cal
and
si
m
ulati
on
are
the
sam
e
with
the
pr
im
ary
sy
stem
par
a
m
et
e
rs.
The
resu
lt
s
can
be
co
ns
ide
red
a
s
the r
ec
omm
en
dation f
or E
H
r
el
ay
ing
n
et
w
ork researc
h.
REFERE
NCE
S
[1]
S.
Bi,
et
al
.
,
“
W
ire
le
ss
powere
d
c
om
m
unic
at
ion:
Opportunit
ie
s
an
d
cha
llenges
,”
IEE
E
Comm
unications
Magazine,
vol/
issue:
53
(4),
pp.
117
-
125
,
20
15
.
[2]
D.
Ni
y
at
o
,
et
al
.
,
“
W
ire
l
ess
Pow
ere
d
Com
m
unic
at
ion
Net
works
:
Resea
rc
h
Dire
c
ti
ons
a
nd
T
e
chnol
ogi
c
a
l
Approac
hes
,”
I
E
EE
W
irel
ess Co
mm
unic
ati
ons,
p
p.
2
-
11
,
2017
.
[3]
A.
A.
Nasir
,
et
al
.
,
“
Re
lay
ing
Protocol
s
for
W
ire
le
ss
Ene
rg
y
Ha
rve
sting
a
nd
Inform
at
ion
Proce
ss
ing
,”
I
E
EE
Tr
ansacti
ons on Wirele
ss
Comm
unic
ati
ons
,
vol
/i
ss
ue:
12
(7)
,
pp
.
36
22
-
3636
,
2013
.
[4]
R.
Atal
l
ah,
e
t
al
.
,
“
Ene
rg
y
har
vesti
ng
in
vehi
cu
la
r
net
w
orks:
A
cont
emporar
y
surve
y
,”
IEEE
Wireless
Comm
unic
ati
ons,
vol
/i
ss
ue:
23
(2)
,
pp
.
70
-
77
,
201
6
.
[5]
H.
Yu,
et
a
l
.
,
“
W
hat
is
5G
?
E
m
erg
ing
5G
Mo
bil
e
Se
rvi
ce
s
an
d
Network
Requi
rement
s
,”
Susta
inabi
lity
,
vol/
iss
ue
:
9
(10),
pp
.
1848
,
2017
.
[6]
K.
Huang
and
V.
K.
La
u
,
“
En
abl
ing
W
ireless
Pow
er
Tra
nsfer
in
Cellula
r
N
etw
orks:
Archi
tect
ure
,
Mode
li
ng
a
nd
Deplo
y
m
en
t
,”
I
EE
E
Tr
ansacti
o
ns on
Wireless
C
omm
unic
ati
ons,
vol/
issue:
13
(2),
pp.
902
-
912
,
20
14
.
[7]
X.
Zhou,
e
t
al
.
,
“
W
ire
le
ss
Inform
at
ion
and
Pow
er
Tra
nsfer
:
Arc
hit
e
ct
ure
Design
and
Rat
e
-
Ene
rg
y
Tr
ade
off
,”
I
E
EE
Tr
ansacti
ons on Com
municat
ions,
vol
/i
ss
ue:
61
(1
1),
pp
.
4754
-
476
7
,
2013
.
[8]
F.
D
.
Rango
an
d
M.
Trope
a
,
“
Ene
rg
y
saving
a
nd
loa
d
balanc
i
ng
in
wire
le
ss
a
d
hoc
net
works
through
ant
-
b
ase
d
routi
ng
,”
P
erfor
mance
Ev
alua
ti
o
n
of
Computer
&
T
el
ec
omm
unic
ati
on
Syste
ms
,
2009.
SPE
CT
S
2009.
Inte
rnation
al
Symposium on
, v
ol.
41
,
pp
.
117
-
1
24
,
2009
.
[9]
F.
D
.
Rango
an
d
M.
Tropea,
“
Sw
arm
int
el
li
g
e
nce
base
d
ene
rg
y
saving
and
lo
ad
balanc
ing
in
wire
le
ss
ad
ho
c
net
works
,”
i
n
P
roce
edi
ngs
of
th
e
2009
worksho
p
on
Bi
o
-
inspired
algorit
hms
for
distribut
ed
systems
,
pp.
77
-
84
,
2009
.
[10]
F.
D
.
Rango,
et
a
l
.
,
“
MD
MC: A
W
SN
coope
rat
iv
e
protoc
o
l
for
M
ini
m
iz
ing
th
e
Da
ta
Distort
ion,
”
P
roce
edi
ngs of
th
e
2018
11th
I
FIP
Wirel
ess and Mo
bil
e
Net
work
ing Confe
renc
e, W
M
NC
,
2018
.
[11]
H.
Ju
and
R
.
Zh
ang,
“
Optimal
R
esourc
e
Al
location
in
Full
-
Dupl
ex
W
ire
l
es
s
-
Pow
ere
d
Com
m
unic
ation
Network
,”
IEE
E
Tr
ansacti
o
ns on
Comm
unications,
vo
l/
issue:
62
(10),
pp.
352
8
-
3540
,
2014
.
[12]
Q.
Yao,
e
t
al
.
,
“A
dapt
ive
h
arv
es
t
-
the
n
-
coope
r
at
e
:
Delay
-
awa
re
w
ire
l
ess
powere
d
comm
unic
at
ion
net
works
,”
2016
IEE
E
17th
Inte
r
nati
onal Wor
kshop
on
Signa
l Proce
ss
ing
Ad
vance
s in
Wire
le
ss
Co
mm
unic
ati
ons (
SPA
WC)
,
2016
.
[13]
Y.
Gu,
et
al
.
,
“
An
ada
pti
ve
tr
an
sm
ission
protoc
ol
for
wire
le
ss
-
pow
e
red
coope
r
ative
comm
unic
at
i
ons
,”
2015
IEEE
Inte
rnational
Co
nfe
renc
e
on
Co
mm
unic
ati
ons (
ICC)
,
2015
.
[14]
H.
Ju
and
R.
Zha
ng,
“
Throu
ghput
Maximiz
at
ion
in
W
irele
ss
Po
were
d
Com
m
unic
at
ion
Networks
,”
IE
E
E
Tr
ansacti
ons on Wirele
ss
Comm
unic
ati
ons
,
vol
/i
ss
ue:
13
(1)
,
pp
.
41
8
-
428
,
2014
.
[15]
K.
Huang
and
V.
K.
La
u
,
“
En
abl
ing
W
ireless
Pow
er
Tra
nsfer
in
Cellula
r
N
etw
orks:
Archi
tect
ure
,
Mode
li
ng
a
nd
Deplo
y
m
en
t
,”
I
EE
E
Tr
ansacti
o
ns on
Wireless
C
omm
unic
ati
ons,
vol/
issue:
13
(2),
pp.
902
-
912
,
20
14
.
[16]
Y.
Ma,
et
al
.
,
“
Distribut
ed
r
esourc
e
allocation
for
power
bea
con
-
a
ss
iste
d
wire
le
ss
-
powere
d
comm
u
nic
a
ti
ons
,”
201
5
IEE
E
Inte
rnat
io
nal
Conf
ere
nce
on
Comm
unic
ations
(
ICC
)
,
2015
.
[17]
J.
Park,
et
al
.
,
“
Ene
rg
y
B
ea
m
form
ing
for
W
ire
les
s
Pow
er
Tra
nsfer
in
MIS
O
Het
ero
gene
ous
Ne
t
work
W
i
th
Pow
er
Bea
con
,”
IEEE C
omm
unic
ati
ons Let
t
ers,
vol
/i
ss
ue:
21
(5)
,
pp
.
116
3
-
1166
,
2017
.
[18]
V.
Inz
i
ll
o,
e
t
a
l
.
,
“
An
Adapt
ive
Bea
m
form
ing
Ti
m
e
with
Rou
nd
-
Robin
MA
C
Algorit
hm
for
Reduc
ing
Ene
r
g
y
Consum
pti
on
in M
AN
ET
,”
Journal
of
Sensor an
d
Actuator
Ne
tw
orks
,
vol/is
sue:
7
(4),
pp
.
50
,
2018
.
APPE
ND
I
X
A
Fr
om
(
15
),
we have:
00
0
0
0
00
1
00
0
(
)
(
)
(
)
(
)
(
)
1
2
2
t
h
t
h
th
th
th
th
Y
k
k
Y
AF
o
u
t
Y
Y
X
k
k
R
D
B
R
th
R
D
B
R
th
YY
th
th
k
P
f
y
d
y
f
y
d
y
f
x
d
x
YY
f
y
d
y
f
y
d
y
K
k
Y
k
Y
(A1)
0
1
00
1
2
2
(
)
th
AF
R
D
B
R
th
R
D
B
R
th
o
u
t
Y
th
th
k
YY
P
K
f
y
dy
k
Y
k
Y
(A2)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
5
,
Oct
ober
20
19
:
3
3
9
9
-
3
4
0
6
3406
So
,
w
e
can
com
pu
te
()
Y
fy
by usi
ng e
qu
at
io
n [
8.4
86,18] i
n [
22
]
1
(
(
)
)
(
)
vv
vv
d
z
K
z
z
K
z
dz
(A3
)
0
()
(
)
2
(
2
)
Y
Y
S
R
B
S
S
R
B
S
d
F
y
f
y
K
y
dy
(A4)
Com
bin
e (A3
),
(A4
),
OP i
s ob
ta
ined
as
in (
16).
APPE
ND
I
X
B
0
0
0
0
1
Pr
,
1
Pr
Pr
DF
t
h
t
h
t
h
t
h
out
P
X
Y
X
Y
k
k
k
k
(B1)
0
0
0
P
r
1
P
r
1
t
h
t
h
t
h
X
X
X
F
k
k
k
(B2)
Fr
om
(
9)
we o
btain:
1
0
0
0
P
r
2
2
t
h
R
D
B
R
t
h
R
D
B
R
t
h
XK
k
k
k
(B3)
Si
m
il
ar as
ab
ove,
we
ca
n
c
om
pu
te
:
1
0
0
0
P
r
2
2
t
h
S
R
B
S
t
h
S
R
B
S
t
h
YK
k
k
k
(B4)
Fr
om
(
B4
),
(
20)
is
dem
on
strat
ed.
Evaluation Warning : The document was created with Spire.PDF for Python.