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.
8
,
No.
6
,
D
ece
m
ber
201
8
, pp.
4374
~
43
81
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
8
i
6
.
pp
4374
-
43
81
4374
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
Wireles
s
Mesh N
etworks
Based on
MBPSO
Al
gorith
m
to
Improve
ment Th
ro
u
ghpu
t
Shiv
an
Qa
sim
A
mee
n
1
,
Fir
as L
ayth Kh
ale
el
2
1
Univer
siti
Keba
ngsaa
n
Mal
a
y
s
ia,
Facu
lty
of
Info
rm
at
ion
Sci
ence
a
nd
T
ec
hnolo
g
y
,
Softam
Depa
r
t
m
ent
,
Mal
a
y
sia
2
Ti
krit Unive
rsi
t
y
,
Facu
lty
of
Co
m
pute
r
Scie
n
ce
Com
puer
Scie
nc
e
Depa
rtment
,
S
al
ah
Din
,
Ira
q
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Dec
24
, 201
7
Re
vised
Ma
r 1
0
, 2
01
8
Accepte
d
Ma
r
24
, 201
8
W
ire
le
ss
Mesh
Networks
ca
n
be
reg
ard
ed
a
s
a
t
y
p
e
of
co
m
m
unic
at
ion
te
chno
log
y
in
m
esh
topol
og
y
in
which
wire
l
ess
nodes
int
er
conn
ec
t
with
on
e
anot
her
.
W
ire
l
es
s
Mesh
Networks
depe
nding
on
t
he
sem
i
-
stat
i
c
co
nfigura
t
ion
in
diff
ere
nt
pat
h
s
among
nodes
such
as
PD
R,
E2
E
de
l
a
y
and
thro
ughput.
Thi
s
stud
y
sum
m
ari
ze
d
diffe
ren
t
t
y
pe
s
of
pre
vious
heur
isti
c
a
lgori
thm
s
in
orde
r
to
ada
pt
with
prop
er
a
lgori
thm
th
a
t
coul
d
solve
th
e
issue.
Th
ere
fo
re,
the
m
ai
n
obje
c
ti
ve
of
thi
s
stud
y
is
to
de
t
ermine
th
e
prop
er
m
et
hods,
app
roa
che
s
o
r
alg
orit
hm
s
th
at
should
be
ada
pt
ed
to
improve
t
he
throughput.
A
Modified
Bina
r
y
Part
ic
l
e
Sw
arm
Optimiza
ti
on
(MBP
SO
)
appr
oa
ch
was
ada
pt
ed
to
improvem
ent
s
t
he
throughpu
t.
Final
l
y
,
th
e
fin
ding
show
s
tha
t
throughpu
t
inc
re
ase
d
b
y
5.
7
9%
from
the previous stud
y
.
Ke
yw
or
d:
Heurist
ic
alg
ori
th
m
MB
PSO
Mi
ni
m
iz
e cost o
f
d
ist
ance
Rou
te
r
s
W
i
reless m
esh netw
orks
Copyright
©
201
8
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
:
Firas Layt
h K
ha
le
el
,
Tikrit U
nive
rsi
ty
-
Fac
ulty
o
f C
om
pu
te
r
Scie
nce,
Com
pu
t
er S
ci
e
nce
Dep
a
rtm
ent,
Sala
h Din
, Ir
aq.
Em
a
il
:
Firas_L
ay
th@tu.
e
du.i
q
,
Firas
_Layt
h@y
ah
oo.co
m
1.
INTROD
U
CTION
A
m
esh
topolog
y
in
w
hich
rad
i
o
no
des
ar
e
arr
a
nged
,
m
akin
g
up
a
c
om
m
un
ic
at
ion
s
netw
ork
a
s
sh
ow
n
in
Fig
ur
e 1
,
is kn
own
a
s a
W
i
reless M
esh Net
w
ork (
WMN
)
[
1]. T
hi
s n
et
w
ork
al
s
o t
akes t
he
f
orm
o
f
an
ad
-
hoc
wireles
s
netw
ork
[
2].
W
M
Ns
usual
l
y
inv
ol
ve
m
es
h
cl
ie
nts,
m
es
h
r
ou
te
r
s,
an
d
gateways.
T
he
m
esh
cl
ie
nts
usual
ly
com
pr
ise
a
va
r
ie
ty
of
wireles
s
ap
p
li
ances
in
cl
ud
in
g
PCs,
ha
ndset
s,
a
nd
t
he
li
ke.
O
n
t
he
oth
e
r
hand,
t
he
m
es
h
r
oute
rs
help
in
traf
fic
f
orwardin
g
to
a
nd
f
ro
m
gatew
ay
s,
w
hich
m
ay
no
t
hav
e
I
nter
net
connecti
on.
T
he
rad
io
node
c
ov
e
ra
ge
area
t
hat
functi
ons
as
on
e
netw
or
k
is
at
tim
es
r
efer
red
t
o
a
s
a
m
esh
cl
oud
a
nd
acce
ssing
this
m
es
h
cl
ou
d
de
pe
nds
wholly
on
the
ra
dio
no
des
functi
onin
g
in
accor
d
with
one
oth
e
r
to
produce
a
r
adio
net
work.
A
m
esh
network
offe
rs
re
dund
a
ncy
an
d
is
reli
able
[1
]
.
Wh
ene
ve
r
a
par
t
ic
ular
node
sto
ps
functi
on
i
ng,
t
he
oth
e
r
no
des
co
ntinu
e
c
om
m
un
ic
at
ing
with
on
e
an
ot
her
directl
y,
via
one
interm
ediat
e
n
od
e
or
m
or
e.
W
i
reless
Me
sh
Net
works
can
sel
f
-
heal
and
sel
f
-
f
orm
[1
]
.
W
irel
ess
Me
s
h
Netw
orks
can
be
exec
uted
vi
a
sever
al
wirel
ess
te
chnolo
gies
com
pr
isi
ng
802.1
6,
80
2.15,
and
80
2.11
c
el
lular
te
chnolo
gies a
nd r
e
quire
no re
stric
ti
on
t
o
a
ny
p
r
oto
c
ol
or
te
chnolo
gy.
The
im
po
rta
nc
e
of
Wireless
m
esh
net
wor
k
le
ads
to
be
us
e
d
i
n
se
ver
al
do
m
ai
ns
su
c
h
a
s
[1
-
10]
.
Also
wireless
m
esh
arc
hitec
ture
i
s
an
i
niti
al
phase
towa
r
ds
offer
in
g
high
dy
nam
ic
-
ba
ndwi
dth
a
nd
c
os
t
-
e
ff
ic
ie
nt
netw
orks
f
or
a
par
ti
cula
r
co
ve
rag
e
a
rea.
E
xcl
ud
i
ng
t
he
cabli
ng
betwee
n
nodes,
a
netw
ork
of
routers
m
akes
up
the
wireless
m
esh
in
fr
ast
r
uct
ur
e
.
This
c
om
pr
ise
s
pee
r
ra
dio
app
li
ance
s
tha
t
req
ui
re
no
wirin
g
to
a
cable
d
port
,
un
li
ke
tr
a
diti
on
al
acce
ss
points
(
AP)
in
W
LAN.
By
se
pa
rati
ng
t
he
distances
int
o
a
s
uc
cessi
on
of
sm
a
ll
hops,
the
m
esh
inf
ras
tructu
re
ca
n
c
onvey
data
t
hro
ugh
la
rg
e
dista
nces.
I
nterm
ediat
e
nodes
help
boost
the
sig
na
l
and
al
so
c
ooper
at
e
in
tra
ns
m
itti
ng
data
f
ro
m
a
pa
rt
ic
ular
po
i
nt
to
a
nothe
r
po
i
nt
(e
.g.,
Poi
nt
A
t
o
P
oin
t
B)
by
m
aking
decisi
on
s
f
or
f
orwardin
g
based
on
their
unde
rstand
i
ng
of
the
ne
twork
,
i.e.
t
hro
ugh
im
ple
m
entin
g
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
& C
om
p
Eng
IS
S
N:
20
88
-
8708
Wi
rel
ess Mesh
Ne
tw
or
ks B
as
e
d on MBP
SO A
lgo
rit
hm t
o
…
(
Sh
iv
an
Q
as
im
Am
ee
n
)
4375
routin
g.
T
his
ty
pe
of
arc
hitec
ture
m
igh
t,
with
ca
utiou
s
desig
n,
offe
r
an
ec
onom
ic
ad
van
ta
ge,
s
pectral
eff
ic
ie
ncy,
and
h
ig
h ba
ndwidt
h
al
l t
hro
ugho
ut
the cove
rag
e
area.
W
i
reless
Me
s
h
Netw
orks
have
a
com
par
at
ively
ste
ady
topolo
gy
apa
rt
fro
m
the
rar
e
m
alf
unct
io
ning
of
their
no
des
or
ad
de
d
-
on
nodes.
T
her
e
are
infr
e
quent
cha
ng
e
s
happe
ning
to
the
traff
ic
path,
as
these
r
esult
s
from
the
agg
r
e
gation
of
a
huge
nu
m
ber
of
e
nd
us
e
rs.
Vi
rtua
ll
y
a
ll
infr
ast
r
uc
ture
m
esh
network
tra
ff
ic
is
ei
ther
forw
a
r
ded
t
o
the
gate
way
or
from
i
t,
wh
il
e
in
cl
ie
nt
m
es
h
net
wor
ks
or
ad
-
hoc
net
wor
ks
the
flo
w
of
traff
ic
occurs am
id ar
bitrary
node pa
irs [4
6].
Figure
1
.
W
irel
ess
m
esh
netw
ork diag
ram
[
2]
This
ki
nd
of
infr
a
struct
ur
e
m
ay
be
centra
ll
y
han
dled
(
usi
ng
a
ce
ntral
serv
e
r)
or
de
centrali
zed
(w
it
ho
ut
any
c
entral
ser
ve
r)
[
47
]
,
[
11]
T
hes
e
ty
pes
are
c
om
par
at
ively
low
-
c
os
t
a
nd
m
ay
be
ve
ry
resil
ie
nt
an
d
reli
able,
as
the
ind
ivi
du
al
node
requires
on
ly
the
transm
ittance
to
the
degr
ee
of
the
nex
t
node.
Nodes
functi
on
as
routers
f
or
t
he
tra
ns
m
issi
o
n
of
da
ta
rangi
ng
from
no
de
s
that
are
cl
o
se
by
to
fa
r
away
peer
s
,
w
hich
cannot
be
reac
hed
in
just
one
hop,
le
adin
g
to
a
network
th
at
m
ay
sp
a
n
longer
dis
ta
nces.
A
m
esh
networ
k
topol
og
y
is
al
so
depen
da
ble,
as
ever
y
no
de
is
coupled
to
a
few
ot
her
nodes
.
On
ce
a
node
fall
s
out
of
the
net
wor
k,
as
a
res
ult
of
t
he
fa
il
ur
e
of
t
he
ha
rdwar
e
or
a
ny
oth
e
r
ca
us
e,
it
s
neig
hbors
wi
ll
swiftly
deter
m
ine
an
al
te
r
na
ti
v
e
route
via a ro
ut
ing
prot
oco
l.
2.
RESEA
R
CH B
AC
K
G
ROU
ND
W
i
reless
Me
sh
Netw
orks
can
be
re
garde
d
as
a
ty
pe
of
com
m
un
ic
at
ion
te
chnolo
gy
in
m
e
sh
to
po
l
ogy
in
wh
ic
h
wirel
ess
nodes
inte
r
connect
with
one
an
oth
e
r
[
12
]
,
[
13
]
.
Me
s
h
netw
ork
com
m
un
ic
at
ion
too
ls
are
com
m
on
ly
gr
oupe
d
as
r
ou
te
rs
,
gate
ways,
a
nd
cl
ie
nts.
I
n
th
es
e
netw
orks
,
ever
y
gateway
m
igh
t
directl
y
offer
a
ty
pe
of
ser
vice,
and
the
data,
wh
ic
h
flo
w
s
a
m
id
gateways
and
subsc
riber
s
,
are
relay
ed
via
router
s.
IEE
E
802.1
1,
I
EEE
802.1
5,
a
nd
IE
EE
802.1
6
are
so
m
e
of
the
te
chnolo
gies
in
wh
ic
h
the
net
work
i
ng
of
wi
reless
m
esh
fin
ds
it
s
app
li
cat
ion
a
nd
a
re
pres
um
ed
to
be
th
e
prov
i
der
ca
r
rier
f
or
t
he
pr
ob
le
m
design.
W
he
n
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
20
88
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
8
, N
o.
6
,
Dece
m
ber
201
8
:
437
4
-
4381
4376
consi
der
i
ng
th
e
su
cces
s
recorde
d
by
W
i
-
Fi
te
ch
no
l
og
y
(IEEE
802.1
1)
in
the
s
ubsti
tuti
on
of
wire
d
ne
twor
k
com
pu
te
rs
fro
m
of
fices
an
d
house
holds
,
t
he
m
ai
n
obj
ec
ti
ve
of
this
re
search
an
d
in
du
st
ry
is
to
ha
rn
es
s
resou
rces
to
wa
rd
s
e
ra
dicat
ing
the
cost
of
se
tt
ing
up
a
nd
m
ai
ntaining
ca
ble
us
e
i
n
m
etr
op
olit
an
broa
db
a
nd
netw
ork
a
reas.
This
researc
h
resu
lt
ed
in
t
he
enh
a
ncem
ent
of
IE
EE
802.1
6,
w
hich
now
pe
rfor
m
s
the
f
unct
i
on
of
a
bac
kh
a
ul
for
br
oadba
nd
wireless
acce
ss
f
or
(
W
M
A
N)
m
et
ro
poli
ta
n
wi
reless
net
work
areas
[
14]
.
The
intero
per
a
ble
a
pp
li
cat
io
n
of
t
he
I
EEE
802.1
6
wireless
fam
il
y
is
po
pula
rly
know
n
as
Wi
MAX
[
15
]
.
T
he
IEEE
Stand
a
r
ds
Boa
rd
init
ia
te
d
a
f
un
ct
io
ning
gro
up
i
n
1999
to
f
or
m
ulate
sta
ndard
s
for
the
W
MAN
broadba
nd.
The
fou
nd
e
d
gro
up
m
ade
avail
able
their
IEEE
802.1
6
init
ia
l
dr
a
ft
in
Febru
ary
2004.
I
n
this
ph
a
se,
the
(S
S)
su
bsc
ri
ber
sta
ti
on
s
a
nd (
BS
) b
ase sta
ti
on
s
ou
gh
t t
o be im
m
o
bile an
d
in
li
ne
-
of
-
si
gh
t,
r
es
pe
ct
ive
ly
[12
]
,
[
16]
.
Anothe
r
substa
ntial
i
m
pr
ov
e
m
ent
occu
r
red
with
the
IEE
E
802.16
e
-
2005
introd
uctio
n,
wh
ic
h
de
al
s
with
com
m
un
ic
at
ion
an
d
m
ob
il
it
y
that
are
non
-
li
ne
-
of
-
sigh
t
f
or
the
su
bsc
ri
ber
s
be
tween
BS
an
d
SS,
util
iz
at
ion
of
Scal
able
O
rth
ogonal
F
reque
nc
y
Divi
sio
n
M
ulti
ple
Access
,
enh
a
nce
d
se
r
vice
qual
it
y
suppo
rt
,
and
m
uch
m
or
e
[16].
Pr
act
ic
al
pro
blem
s
still
exist
eve
n
w
it
h
the
le
vel
of
de
velo
pm
ent
recorde
d
su
c
h
as
the
requirem
ent
for
acce
ss
of
uneven
t
raffic
di
stribu
ti
on
in
de
ns
el
y
popula
te
d
areas
,
the
s
ign
al
-
to
-
no
ise
rati
o
occurri
ng
at
th
e
edg
e
of
the
cel
l,
and
c
over
age
hole
s
that
e
m
erg
e
as
a
re
su
lt
of
non
-
li
ght
-
of
-
sig
ht
net
works
and
s
ha
dowing,
et
c.
T
he
W
i
MAX
protoc
ols
hav
e
to
guar
antee
de
penda
bili
ty
,
add
ress
cov
e
ra
ge
hole
s
,
an
d
su
pp
or
t
utm
os
t
m
ob
il
it
y
to
com
p
et
e
with
wir
ed
broa
db
a
nd
pro
vid
er
s
a
nd
3G.
Eac
h
c
halle
ng
e
is
in
co
nt
rast
to
the
oth
e
r.
Re
li
abili
ty
decr
eases
as
rate
of
da
ta
increases.
Howe
ver,
coverage
area
(i
.e.
the
siz
e
of
the
cel
l)
reduces
as
th
e
reli
abili
ty
of
s
erv
ic
e
i
ncr
eas
e
s.
Re
duct
io
n
of
the
siz
e
of
th
e
cel
l
w
ou
l
d
re
su
lt
in
a
n
incr
e
ase
in
the qua
ntit
y of
BSs f
or a s
peci
fic area
co
ver
a
ge,
w
hich wil
l resu
lt
in
the
r
is
e of
netw
ork
c
os
ts
[12
]
,
[
16]
.
Re
la
y
station
(RS)
insertio
n
be
tween
the
BS
s
and
SSs
ser
ve
s
as
the
best
s
olu
ti
on
at
this
po
i
nt,
w
hic
h
will
ro
ute
data
in
betwee
n
th
e
sta
ti
on
s.
T
he
relay
is
util
iz
e
d
f
or
the
e
xten
sion
of
netw
ork
co
ver
a
ge
ra
nge
an
d
capaci
ty
,
wh
ic
h
al
so
co
nn
ect
s
the
cov
e
rag
e
ho
le
s
e.
g.
,
s
ha
dows
of
buil
di
ng
s
,
there
fore,
enh
a
ncin
g
en
d
-
to
-
e
nd
com
m
un
ic
at
ion
q
ualit
y
[
17]
.
IEEE
80
2
.
16j
i
s
a
m
od
ifie
d
ve
rsion o
f
IE
EE
802.1
6e.
It
pr
oj
ect
s
that data
am
id
a
SS
an
d
BS
can
be
relay
ed
t
hro
ugh
a
RS
via
MM
R
(
m
ob
il
e
m
ulti
-
ho
p
relay
netwo
r
k),
w
hich
util
iz
es
the
stren
gth
s
of
wireless
m
ulti
-
hop
c
onnecti
vity
[18].
T
he
rang
e
of
c
ov
e
rag
e
and
qua
ntit
y
of
W
im
ax
is
ex
pa
nd
e
d
with
the
intr
oductio
n
of
I
E
EE
80
2.16j,
a
ddressi
ng
the
pro
blem
of
bu
il
di
ng
co
ve
r
age
ho
le
s
,
a
nd
th
us
si
m
plifyi
ng
th
e
extensio
n
of
cov
e
ra
ge
tem
po
ra
rily
to
areas
with
a
high
-
densi
ty
po
pula
ti
on
.
T
he
arc
hitec
ture
of
the
netw
ork
pr
ese
nts
se
ve
r
al
co
m
plica
ti
on
s
withi
n
the
pr
e
viously
cha
ll
eng
ed
rad
i
o
acce
ss
netw
orks
that
pro
vid
e
s
uppor
t
fo
r
m
ob
il
it
y
[19
]
,
[
20]
e.
g.
,
t
he
sc
hedulin
g
of
c
ha
nn
el
acc
ess
regar
ding
f
reque
ncy
reuse
,
RS
and
BS
place
m
ent,
resour
c
e
(ti
m
e
and
f
reque
ncy)
al
locat
ion
,
f
reque
n
cy
and
ti
m
e,
et
c.
[12
]
,
[
16]
.
The
su
bse
que
nt
s
ubsect
ion
s
pro
vide
sam
ples
of
va
rio
us
researc
he
s
car
ried
out
on
the
net
works
of
IEEE
80
2.16j
t
o
dev
el
op
a
grea
te
r
le
vel
of
inf
or
m
at
ion
reg
a
r
ding
netw
orks
and
pro
blem
s
encoun
te
red
i
n
wireless
ne
twor
k
desig
n, w
hi
ch
sh
are
s
om
e si
m
il
arity to th
e
pro
blem
h
igh
li
gh
te
d
in
this st
ud
y.
In
wireless
co
m
m
un
ic
at
ion
,
interfe
re
nce
m
a
y
arise
du
e
to
s
har
i
ng
of
c
omm
un
ic
at
ion
m
e
diu
m
withi
n
the
sta
ti
on
s
.
T
o
s
olv
e
t
his
pro
blem
,
network
res
ources
[
21
]
s
uggest
a
sche
du
li
ng
al
gorithm
,
w
hic
h
will
su
pp
or
t
s
patia
l
reu
se
gains
f
r
om
the
two
hops
in
a
net
wor
k
that
is
relay
-
enab
le
d.
I
brahi
m
et
.
al.
and
Ge
e
t.
al.
disco
ver
e
d
t
hat
durin
g
thei
r
st
ud
y
of
a
n
a
nal
yt
ic
al
m
od
el
to
exam
ine
the
capaci
ty
of
a
ce
ll
with
the
exte
ns
io
n
of
a
tw
o
-
hop
cov
e
ra
ge,
t
hat
sp
at
ia
l
re
us
e
co
uld
re
du
ce
losses
i
n
ca
pa
ci
ty
[22
]
,
[
23]
.
Mo
re
ov
e
r,
m
any
researc
hes
ha
ve
f
ocu
se
d
on
wireless
netw
ork
locat
io
nal
de
sign.
Also
,
a
good
am
ou
nt
of
resea
rch
ha
s
bee
n
done
in
volvin
g
va
rio
us
wi
re
le
ss
netw
ork
pro
vid
e
rs
an
d
t
he
to
po
l
og
ic
al
and
a
rch
it
ect
ur
al
pla
nn
i
ng
for
the
netw
orks.
[
24
]
,
[
25
]
use
d
a
n
integer
pro
gr
a
m
m
ing
m
od
el
includi
ng
m
any
al
go
rithm
s
ba
sed
on
T
ab
u
Searc
h
and
Gr
ee
dy
[26]
to
decide
be
tt
er
po
sit
ions
to
cov
e
r
va
rio
us
traf
fic
con
c
entrati
ons
us
i
ng
BSs
[24
]
,
[
25]
.
The
m
ajo
r
c
halle
nge
face
d
by
the
m
ob
il
e
ind
us
try
is
the
m
igrati
on
from
2G
to
3G
net
works
in
a
w
ay
that
sat
isfact
ion
of
the
cu
stom
er
is
achie
ved
wit
h
a
l
ow
co
st
of
operati
on,
an
d
the
nu
m
ber
of
cel
ls
that
is
us
ed
is
al
so
reduce
d
to
the
m
ini
m
u
m
.
Kaur
et
.
al
.
propose
d
t
he
det
erm
inati
on
of
cel
l
sit
es
us
in
g
a
heurist
ic
al
gorithm
,
wh
ic
h
works
by
ra
nk
i
ng
cel
ls
from
the
ge
ner
at
e
d
sim
ulate
d
data
a
nd
c
el
l
rem
ov
al
from
the
per
i
odic
al
l
y
si
m
ulate
d
m
od
el
[27]
.
A
pr
el
im
inary
stud
y was con
du
ct
e
d
by
Si
nha
et
. a
l
.
a
nd
D
opp
le
r
et
. a
l
.
in r
ega
rd
to
the IEE
E 8
02.
16
j
desig
n
str
uctu
r
e
[28
]
,
[
29]
.
O
ne
pre
vious
w
ork
de
sig
ned
a
program
m
ing
m
od
el
fo
r
inte
ger
s
on
IEE
E
802.1
6d
netw
orks
an
d
su
ggest
e
d
so
lv
ing
the
issues
of
creati
ng
the
lowest
co
st
backhau
l
wir
el
ess
network
us
i
ng
heurist
ic
al
gor
it
h
m
s
(w
it
h
BSs)
t
o
fu
l
fill
the
S
S
require
m
ents
with
no
effe
ct
on
t
he
capaci
ty
lim
i
ts
in
BS
s
[30]
.
A
num
ber
of
case
st
udie
s
ha
ve
bee
n
prese
nted
t
o
determ
ine
relay
op
ti
m
al
locat
ion
a
nd
t
o
push
the
syst
e
m
ou
tpu
t
to
the
m
ax
within
the
B
S
-
RS
cel
l
cov
er
a
ge
in
802.1
6j
net
works
[
22
]
.
H
ow
e
ve
r,
these
works
fail
to
ta
lk a
bo
ut the iss
ues
as
so
ci
at
ed wit
h m
ul
ti
ple
-
relay
p
la
nnin
g.
Anothe
r
stu
dy
dev
el
op
e
d
he
uri
sti
c
-
base
d
RS
s
to
ad
dress
th
e
issue
of
netw
ork
dep
l
oym
en
t
and
reuse
of
r
adi
o
re
souc
e
wh
e
n
IEEE
802.1
6j
MM
R
ne
tworks
a
re
i
nvol
ved
[
31
]
.
Y
et
ano
t
her
st
udy
assessed
t
he
abili
ty
of
a
netw
ork
f
or
IE
EE
802.1
6j
via
co
opera
ti
ve
di
ver
sit
y
f
or
upli
nk
tra
nsm
issi
on
s
[32].
The
res
ult
f
rom
the
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
& C
om
p
Eng
IS
S
N:
20
88
-
8708
Wi
rel
ess Mesh
Ne
tw
or
ks B
as
e
d on MBP
SO A
lgo
rit
hm t
o
…
(
Sh
iv
an
Q
as
im
Am
ee
n
)
4377
researc
h
can
be
e
m
plo
ye
d
in
analy
zi
ng
the
pros
an
d
co
ns
of
capaci
ty
im
pr
ov
em
ent
relay
and
de
ploym
en
t
cost.
Vasishta
et
.
al
.
discusse
d
a
pr
ogram
m
ing
f
or
m
ula
ti
on
f
or
in
te
ger
s
i
n
po
sit
ion
i
ng
BS
a
nd
RS
with
t
he
aim
of
re
duci
ng
the
c
os
t
of
e
sta
blishm
ent
un
de
r
the
lim
it
a
ti
on
of
us
e
r
tra
ff
ic
dem
and
[
33]
.
Bo
und
te
c
hniq
ues
and
a
sta
ndar
d
br
a
nc
h
we
re
app
li
ed
in
s
ol
ving
this
pro
ble
m
.
Howe
ver
,
un
li
ke
the
cas
e
of
sm
all
instances,
wh
e
n
it
com
e
s
to
so
lvin
g
m
et
ro
poli
ta
n
-
s
cal
e
la
rg
e
instances,
this
a
ppr
oac
h
is
lim
i
te
d.
I
n
one
stud
y,
a
cl
us
te
rin
g
a
ppr
oach
was
us
e
d
on
a
sim
i
la
r
pro
blem
[3
3].
To
ac
hieve
the
le
ast
am
ou
nt
of
RSs
i
n
a
n
MM
R
netw
ork,
Chang
et
. a
l
.
rec
omm
end
e
d
a
h
e
uri
sti
c algori
thm
[34
]
.
A
m
od
el
for
relay
-
centric
hi
erarch
al
opti
m
iz
at
ion
,
w
hich
can
be
us
e
d
f
or
both
op
tim
iz
at
ion
of
netw
ork
plan
ni
ng
f
or
RSs
a
nd
ra
dio
r
es
ource
MM
R
netw
orks
m
anag
em
ent,
was
al
so
propo
s
ed
[
35
]
.
Th
e
aim
s
of
t
he
researc
h
are
to
m
ake
the
m
os
t
of
the
util
iz
at
ion
of
t
he
RSs
an
d
ac
hieve
optim
u
m
reserve
d
band
width.
In
a
new
al
go
r
it
h
m
form
ulatio
n
of
t
he
opti
m
iz
at
ion
issue,
the
pro
blem
s
of
assig
nm
ent
const
raine
d
by
cha
nce
are foc
us
e
d on,
so
as to
achie
ve
the ide
al
d
eci
sion
s
on rel
ay
po
sit
io
ning a
nd
base s
ta
ti
on s
el
ect
ion
.
Shim
et
.
al
.
produce
d
a
pa
pe
r
on
t
he
us
e
of
IE
EE
802.1
6j
te
ch
no
l
ogy
for
the
en
ha
ncem
ent
of
infr
a
struct
ur
e
c
omm
un
ic
at
ion
in
ve
hicula
r
wi
reless
net
wor
ks
[
36
]
.
T
hey
presum
ed
the
ve
hicular
SSs
l
oc
at
ion
as
kn
own,
an
d
us
e
d
the
detai
l
ob
ta
ine
d
f
or
th
e
ideal
placem
ent
of
RS
s
in
a
way
that
the
e
nd
-
to
-
en
d
ca
pa
bili
ty
was
ta
ke
n
f
ull
adv
a
ntage
of.
The
stu
dy
c
ombine
d
a
m
od
el
of
high
way
m
ob
il
it
y
and
a
nonlinea
r
op
ti
m
iz
at
ion
m
od
el
f
or the
problem
. Th
e m
od
el
so
l
ution g
uar
a
ntees
ultim
at
e end
-
to
-
e
nd
capab
il
it
ie
s for SSs.
3.
THE
ADVA
N
TAGES
A
ND
DISAD
V
AN
T
AGES OF
PR
EVIO
US
AL
GORIT
HM
S
In
Ta
ble
1,
a
s
umm
ary
of
dif
fer
e
nt
al
gorith
m
s
based
on
th
e
adv
a
ntage
a
nd
disa
dv
a
ntage
of
each
one
is
pr
ese
nted
.
Fi
nally
,
the
previ
ou
s
sect
ion
sum
m
arizes
the
al
gorithm
s,
m
eth
ods
a
nd
ap
pro
aches
in
t
his
fi
el
d
as
sh
ow
n
in
Tabl
e
1.
A
t
the
sam
e
tim
e,
ideal
le
arn
i
ng
e
nvir
onm
ents
can b
e
c
reated.
I
n
oth
er
w
ords,
t
he
He
ur
is
ti
c
al
gorithm
fo
r
the
W
irel
ess
M
esh
Net
work
ha
s
to
ada
pt
the
PSO
a
ppro
a
ch
beca
us
e
thi
s
approac
h
is
m
or
e
com
pr
ehe
ns
ive
than
ot
her
al
gorithm
functi
onal
it
ie
s.
Also,
a
wide
ra
nge
of
c
on
ti
nu
ou
s
optim
iz
at
ion
prob
le
m
s
can
be
a
ddress
ed via t
he
su
cc
essfu
l
appli
cat
i
on of PSO
.
Table
1
.
Partic
le
Sw
a
rm
O
ptim
iz
at
ion
Para
m
et
ers
and
Ch
os
e
n
V
al
ues
No
.
Alg
o
rith
m
Ad
v
an
tag
es
Disad
v
an
tag
es
1
Heu
ristic
Alg
o
rith
m
[
3
7
,
38]
The p
erfo
r
m
an
ce
of
the p
rop
o
sed
sch
e
m
e is
no
t
o
n
ly
clos
e to th
e o
p
ti
m
al
m
u
lt
i
-
p
ath
s
o
lu
tio
n
,
it
also
o
u
tp
erfo
r
m
s e
x
istin
g
m
u
lti
-
p
ath
rou
tin
g
sch
e
m
es.
It
d
o
es n
o
t f
o
llo
w a
stan
d
ard
m
ath
e
m
atical
m
o
d
el
.
Lower c
ap
ab
ility
f
o
r
g
en
eraliza
tio
n
.
2
Gen
etic
Alg
o
rith
m
[
3
9
,
40]
Every rou
tin
g
sess
io
n
con
cu
rr
en
cy
tr
a
n
s
m
iss
io
n
is ef
f
ectiv
ely
m
ax
i
m
i
zed th
rou
g
h
elim
in
atio
n
of
in
terferen
ce betwe
en
wireless
m
esh
r
o
u
ters,
u
sin
g
this
algo
rith
m
.
To so
lv
e the p
rob
le
m
of
m
esh
rou
ter
n
o
d
e
p
lace
m
en
t,
Tabu
S
earc
h
,
an
ex
a
m
p
le
o
f
a
lo
cal
search
m
eth
o
d
,
an
d
Genetic algo
rith
m
s
,
wh
ich
ar
e
p
o
p
u
latio
n
-
b
ased
m
e
th
o
d
s,
m
u
st b
e
h
y
b
ridized
.
3
Ad
ap
tiv
e M
ix
ed
Bias
(
A
MB)
Alg
o
rith
m
[
4
1
]
Better perf
o
r
m
an
c
e is ob
serv
ed
with
th
e us
e of
th
e pro
p
o
sed
app
ro
ach co
n
cernin
g
Adap
tiv
e
Mixed
Bias co
m
p
a
red to
bo
th
exis
tin
g
m
ix
ed
b
ias ap
p
roach
es an
d
I
EE
E
8
0
2
.11
M
AC.
Varied pack
et
rate
s, div
erse netwo
rk
to
p
o
lo
g
ies,
an
d
nu
m
erou
s so
u
r
ces th
rou
g
h
exp
erim
e
n
ts
m
u
st
b
e us
ed
to ach
iev
e the chan
g
es in
perf
o
r
m
an
ce a
n
d
in
enh
an
cin
g
r
o
b
u
s
t so
lu
tio
n
s
It
is v
ital to stu
d
y
t
h
e para
m
e
ters o
f
T
ab
u
Sear
ch
f
o
r
m
o
re
in
f
o
r
m
ati
o
n
to en
h
an
ce the Adap
tiv
e
Mixed
Bias
m
eth
o
d
perf
o
r
m
an
ce
.
4
Gree
d
y
Algo
rith
m
Gree
d
y
Algo
rith
m
s
m
o
stly
(
b
u
t no
t a
lway
s)
f
ail
to
f
in
d
globally
o
p
ti
m
al so
lu
tio
n
s, becau
se th
ey
u
su
ally
do
no
t op
e
rate
ex
h
au
stiv
ely
o
n
all
th
e
d
ata.
This
m
eth
o
d
can re
su
lt in th
e pr
ev
en
tio
n
o
f
arr
iv
in
g
at
th
e bes
t
ov
erall
so
lu
tio
n
in th
e f
u
tu
re,
as th
is
m
eth
o
d
can
co
m
m
it
too
ear
l
y
t
o
cer
tain
ch
o
ices.
Fo
r
in
stan
ce,
th
e
G
reedy Alg
o
rith
m
s,
wh
ich
are
co
n
sid
ered
Gree
d
y
ty
p
icall
y
,
f
ail to d
isco
v
er
th
e
p
rob
le
m
of
grap
h
colo
ring
and
that o
f
th
e glo
b
ally
o
p
ti
m
u
m
,
bu
t oth
er
NP
-
co
m
p
lete
pro
b
le
m
s
p
rov
id
e a
so
lu
tio
n
.
Nevertheles
s,
they
f
u
n
ctio
n
ex
h
au
stiv
ely
b
ecaus
e they
a
re
s
wif
t in reachin
g
o
p
ti
m
u
m
ap
p
rox
im
a
tio
n
s.
5
Local sear
ch
alg
o
rith
m
Gen
erall
y
,
a
ll loca
l
searc
h
algo
rith
m
s
y
ield
resu
lts th
at ar
e
bett
er
th
an
the G
reedy
Alg
o
rith
m
s
.
Ho
wev
er,
the
draw
b
ack o
f
this
i
m
p
ro
v
e
m
en
t is a
lo
n
g
er
run
n
in
g
ti
m
e.
6
Variable
Neig
h
b
o
rho
o
d
Search
Alg
o
rith
m
s
Local sear
ch
pro
ce
d
u
re,
in d
eter
m
in
in
g
the
so
lu
tio
n
s to
di
ff
erent
o
p
ti
m
iz
atio
n
pro
b
le
m
s
,
is
v
ery
ef
f
ectiv
e.
Ho
wev
er,
it
can ge
t stu
ck
in a local
m
in
i
m
a.
7
Particle
Swa
r
m
Op
ti
m
izatio
n
W
id
e
-
rang
in
g
pro
b
le
m
s o
f
op
ti
m
izati
o
n
that are
co
n
tin
u
o
u
s can
app
ly
PSO
and
ob
tain
su
ccessf
u
l
resu
lts.
Rarely
in so
lv
in
g
the iss
u
es, do
th
e
m
esh
r
o
u
ter
n
o
d
es f
ace
p
lace
m
en
t pro
b
le
m
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
20
88
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
8
, N
o.
6
,
Dece
m
ber
201
8
:
437
4
-
4381
4378
4.
RESEA
R
CH DESIG
N
The
c
urren
t
re
search
desi
gn
has
four
phase
s
of
m
et
ho
d
ol
og
y
i
n
relat
io
n
to
this
st
ud
y
t
he
An
al
ysi
s
ph
a
se,
Com
pilat
ion
phase,
Innovat
io
n
phase
,
an
d
Vali
datio
n
phase
are
s
how
n.
At
each
ph
a
se,
the
re
ar
e
ste
ps
to
be
c
on
cl
ude
d
be
f
or
e
the
suc
ceedin
g
phase
can
pr
oceed.
Fo
r
exam
ple,
the
A
naly
sis
ph
ase
entai
ls
the
sta
ges
of
a
naly
sis
in
t
his
stu
dy
wh
il
e
the
I
nnovat
io
n
ph
ase
entai
ls
the
e
nh
a
ncem
ent
an
d
desi
gn
phases,
an
d
finall
y,
the
Vali
datio
n
ph
a
se
en
ta
il
s
the
evaluati
on
and
a
pp
li
cat
io
n
sta
ges.
T
he
f
ollow
i
ng
sect
i
on
descr
i
bes
in
detai
l
each
of these
sta
ges.
4.1.
Analysis
Pha
se
This
m
et
ho
d
us
es
t
he
re
su
l
ts
from
pr
evi
ou
s
resea
rc
h
t
o
asses
s
the
current
pro
ble
m
.
Diff
ere
nt
researc
h
ap
pro
aches,
a
nd
the r
esult
qual
it
y
i
s
acqu
ire
d
to
pro
vid
e
a
so
l
ution
to:
1)
the
pr
ob
le
m
of
Un
s
pl
it
ta
ble
Flow
;
2)
the
prob
le
m
of
Bi
n
Packin
g;
3)
t
he
pro
ble
m
of
Ca
pacit
at
ed
Set
C
ov
e
rin
g;
an
d
4)
the
Prob
le
m
of
Set
Cov
e
rin
g.
4.2.
C
ompari
s
on
Phase
This
ph
as
e
in
volves
a
com
pa
rati
ve
stu
dy
to
determ
ine
the
best
m
et
ho
d
t
o
be
a
pp
li
ed
in
this
w
ork.
This
is
done
t
hro
ugh
i
niti
at
ing
a
c
om
par
at
iv
e
stud
y
betwee
n
the
previ
ous
al
gorithm
s
and
the
c
urren
t
m
et
hods
us
e
d
in
this st
udy.
4.3.
I
nn
ovati
on Ph
as
e
The
i
nnovat
ion p
hase c
onside
rs
the
pr
ocess of
the
pro
pose
d
MB
PS
O
al
gorithm
, as
fo
ll
ows:
a.
Partic
le
Sw
a
rm
Optim
iz
at
ion
(PSO
)
b.
Bi
nar
y Pa
rtic
le
Sw
a
rm
O
ptim
iz
at
ion
(
BP
SO)
c.
Mod
if
ie
d
Bi
na
ry Par
ti
cl
e S
wa
rm
O
pti
m
iz
ation
(MBPSO
)
4.4.
E
valua
tio
n
Ph
as
e
To
en
sure
that
the
al
gorithm
works
co
rr
ect
l
y,
the
validat
io
n
phase
go
es
t
hro
ugh
th
ree
m
et
rics
after
the
He
ur
ist
ic
al
gorithm
is
co
m
ple
te
d.
The
s
e
m
et
rics
are
thr
ough
pu
t,
E
nd
-
to
-
E
nd
Dela
y
(
E2E
DEL
A
Y)
,
a
nd
Packet
Deli
very
Ra
ti
o
(P
DR
).
Finall
y,
the
re
su
lt
is
com
par
e
d
wit
h
the
nea
r
est
te
chn
ic
al
st
ud
y
[42]
in
ord
er
to
m
easur
e the
r
e
su
lt
s im
pr
ov
e
m
ent f
r
om
the o
ri
gin
al
on
e
.
5.
PERFO
R
MANC
E
METR
I
CS
Tw
o
basic
pe
rfor
m
ance
m
et
ri
cs
su
ch
as
the
on
e
s
that
run
vi
a
E2E
delay
and
pack
et
deli
ver
y
f
racti
on
hav
e
be
en
pro
po
s
ed
in
num
e
rous
w
orks
[
43
]
,
[
44]
.
Additi
on
al
ly
,
sim
ula
ti
on
is
c
onside
red
with
the
m
ob
il
it
y
patte
rn
of
node
s.
To
achie
ve
delay
and
pac
ket
drop,
Mirjalil
i
et
.
al
.
pr
opos
e
the
us
e
of
a
ran
dom
waypo
int
m
ob
il
i
ty
m
od
el
[
45
]
.
5.1.
Pac
ket De
li
very
Ratio
(PDR
)
The
nu
m
ber
o
f
delivere
d
pac
ke
ts
is
div
ide
d
by
the
destinat
io
n
to g
ive PD
R.
To
cal
culat
e
P
DR
an
d
t
o
determ
ine
the
loss
rate
of
the
pack
et
,
the
nu
m
ber
of
pack
e
ts
giv
en
by
the
app
li
cat
ion
la
ye
r
of
the
sou
r
ce
is
us
e
d
to
div
ide
the
nu
m
ber
of
pac
kets
rec
ei
ved
by
the
destinat
io
n.
I
n
this
way,
the
m
axi
m
u
m
network
thr
oughput
be
com
es
l
i
m
it
ed.
In
the
routin
g
pr
ot
oco
l
,
an
im
per
at
ive
factor
to
be
acc
om
pl
ished
is
PD
R,
as
there is
no m
a
rg
in
f
or
e
rro
r
in
a r
eal
-
li
fe
envi
ronm
ent li
ke
fl
ooding a
nd ea
r
thquake
s.
5.2.
En
d
-
to
-
E
nd Del
ay
(E2E
D
EL
A
Y)
The
data
packet
will
arr
ive
at
the
endpo
i
nt
within
the
ti
m
e
that
is
aver
age
d
out.
F
or
the
m
et
ric
cal
culat
ion
,
th
e
arr
ival
tim
e
of
the
first
dat
a
pac
ket
is
us
e
d
to
subtract
the
tim
e
at
wh
ic
h
the
first
pac
ket
wa
s
transm
itted.
5.3.
Th
ro
ug
h
p
ut
The
ave
ra
ge
ra
ti
o
of
the
t
otal
si
m
ulati
on
tim
e
durati
on
t
o
the
su
cce
ssf
ul
data
pac
kets
is
the
aver
a
ge
thr
oughput
m
e
tric
.
The
unit
of
Kilo
bits
pe
r
sec
ond
(Kby
te
s/se
c)
is
us
e
d
to
m
easur
e
aver
a
ge
th
r
oughput
,
wh
e
re
the
e
ff
i
ci
ency
and
e
ffec
ti
ven
ess
of
t
he
r
outi
ng
pro
tocol
in
receiv
ing
data
pac
ke
ts
by
destinat
ion
i
s
m
easur
ed
.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
& C
om
p
Eng
IS
S
N:
20
88
-
8708
Wi
rel
ess Mesh
Ne
tw
or
ks B
as
e
d on MBP
SO A
lgo
rit
hm t
o
…
(
Sh
iv
an
Q
as
im
Am
ee
n
)
4379
6.
EVAL
UA
TI
O
N RESULT
Cost
opti
m
iz
ation
is
a
n
a
rea
of
resea
rch
in
W
M
Ns.
O
ur
resea
rc
h
is
base
d
up
on
im
pr
ov
in
g
th
e
so
luti
on
propo
sed
by
[42]
K
haled
a
nd
S
ha
h
Mosta
fa
[
42]
.
The
a
utho
r
c
on
si
der
e
d
c
os
t
op
ti
m
iz
ation
without
ta
kin
g
the
dis
ta
nce
betwee
n
nodes
into
c
on
si
der
at
io
n.
Our
resea
rch
therefo
re
upda
te
d
the
op
ti
m
i
zat
ion
functi
on
to
ta
ke
i
nto
c
onsid
erati
on
the
dis
ta
nces
betwee
n
t
he
diff
e
re
nt
node
s,
us
i
ng
the
Mo
difie
d
Bi
nar
y
Partic
le
Sw
a
r
m
Op
tim
iz
a
tio
n
(MBP
SO)
ap
proac
h.
T
he
res
ults
are
posit
ive
a
nd
our
ap
proac
h
s
hows
no
ti
ceable
im
pr
ovem
ent
co
m
par
e
d
to
the
be
nch
m
ark
stu
dy
.
The
PD
R
s
hows
an
a
ppr
ox
im
at
e
incre
a
se
of
22.47%
,
w
he
r
eas
the
E
2E
delay
saw
a
n
ap
prox
im
at
e
decr
ease
of
21.
14%,
a
nd
fi
nally
the
th
rough
pu
t
increase
d by 5.
79% fr
om
the p
re
vious
w
ork a
s sho
wn in Fi
gure
2.
Figure
2
.
Th
r
ough
pu
t c
om
pari
so
ns bet
ween
the origi
nal and m
od
ifie
d o
bj
ect
ive fun
ct
io
n
7.
DISCU
SSI
ON A
ND CON
C
LUSIO
N
The
m
ai
n
con
trib
ution
of
t
his
stud
y
is
the
pr
opos
e
d
pr
op
e
r
MB
PSO
al
gori
thm
s
us
ed
in
this
stu
dy
to
so
lve
t
he
resea
rch
pro
blem
.
T
his
ap
proac
h
is
adap
te
d
f
r
om
pr
e
vious
wor
k,
so
the
m
ai
n
point
of
this
st
udy
is
to
com
par
e
the
resu
lt
of
this
stud
y
with
previo
us
w
ork.
This
w
ork
ass
um
es
the
sa
m
e
cov
era
ge
area
fo
r
al
l
nodes
.
Further
dev
el
op
m
ent
cou
l
d
be
do
ne
by
con
side
rin
g
diff
e
re
nt
cov
e
r
age
areas
acco
rd
i
ng
to
n
ode
energy
and
node
pri
ori
ty
.
In
ad
diti
on,
te
sti
ng
the
pe
rfor
m
ance
on
a
range
of
node
sp
ee
ds
w
ould
be
a
m
or
e
pract
ic
al
scenari
o,
with
consi
der
i
ng
th
e
sp
eed
as
an
i
nf
l
uen
ce
in
the
sta
nd
ar
d
f
unct
ion
.
F
urt
her
m
or
e,
so
m
e
no
de
s
in
the
netw
ork
m
igh
t
no
t
be
t
ru
ste
d
t
o
tra
nsport t
he
pack
et
.
These
no
des
ha
ve
a
natur
e
of
sel
fishness,
s
o
there
sho
uld
be
a
crit
erion
to
detect
and
avo
i
d
them
.
Fu
tu
re
w
orks
s
hould
c
over
th
e
aspects
of
di
ff
e
ren
t
ra
nges
of
c
ov
e
ra
ge
zo
ne,
s
peeds,
a
nd
co
nf
i
de
nce
f
or
th
e
nodes
of
the
netw
ork.
BPS
O
is
sim
ple
and
hi
gh
ly
rob
us
t.
T
he
BP
S
O
ca
n
s
olv
e
m
ul
ti
di
m
ension
al
an
d
m
ul
tim
od
al
op
tim
iz
at
ion
prob
le
m
s
because
this
al
go
ri
thm
us
es
cont
ro
l
par
am
et
ers
that
are
sim
ple.
Mult
idi
m
ensio
nal
f
unct
ion
al
optim
iz
at
ion
pro
blem
s
can
al
s
o
be
a
ddress
ed
with
the
use
of
th
e
BPS
O
al
gorithm
. H
oweve
r, t
his iss
ue
is
beyo
nd th
e sco
pe of
the
current
stu
dy.
Ther
e
are
se
ve
ral
di
recti
ons
i
n
fu
t
ur
e
w
ork
that
co
uld
be
i
m
ple
m
ented
to
en
ha
nce
the
pe
rfor
m
ance
m
et
rics
us
ed
i
n
this
stu
dy.
F
or
i
ns
ta
nce
,
it
is
possible
t
o
c
on
si
der
ot
her
ne
twork
m
et
rics
su
c
h
as
pac
ke
t
siz
e,
nu
m
ber
of
no
de
s,
an
d
sim
ulati
on
ti
m
e.
This
can
inc
rease
t
he
pr
act
ic
al
it
y
and
ef
fici
ency
of
the
al
gorithm
for
functi
onin
g
in
the
real
w
or
l
d.
O
n
t
he
oth
e
r
hand,
f
uture
w
orks
on
t
he
M
BPSO
al
gorith
m
sh
ou
ld
probably
con
ce
nt
rate
on
a
pa
rtic
ular
pa
rt
of
the
MB
PSO
al
gorithm
,
w
hich
co
uld
be
e
xp
ect
e
d
to
pro
vid
e
good
cost
m
ini
m
iz
at
ion
a
m
on
g
no
de
c
onnecti
ons
.
The
researc
h
pro
blem
is
add
r
essed
us
in
g
he
ur
ist
ic
s
that
ar
e
base
d
on
dif
f
eren
t
var
ia
ti
on
s
of
he
ur
ist
ic
al
gorithm
s.
This
stu
dy
te
ste
d
the
al
gorithm
s
base
d
on
500
-
5000
-
siz
ed
node
s.
Finall
y,
this
stud
y
c
om
par
ed
the
al
gorithm
resu
lt
s
ob
ta
ine
d
wi
th
previ
ou
s
w
ork
.
The
fin
ding
s
hows
that
P
DR
sho
ws
a
n
appr
ox
im
at
e
in
creas
e
of
22.
47%,
w
her
eas
the
E
2E
delay
sa
w
a
n
a
ppr
ox
im
at
e
dec
rease
of
21.
14%,
a
nd
finall
y
the
th
r
ough
pu
t
increase
d by 5.
79% fr
om
the b
enc
hm
ark
stu
dy.
8.
CONCL
US
I
O
N
Cost
op
ti
m
iz
ation
is
an
a
rea
of
resea
rc
h
with
re
gard
to
W
M
Ns.
T
his
researc
h
was
based
upon
i
m
pr
ovin
g
the
so
luti
on
propose
d
by
Nle
y
a
and
Sindiso
M.
[
15
]
in
t
his
re
ga
rd.
T
hey
hav
e
c
onside
r
ed
co
st
op
ti
m
iz
ation
w
it
ho
ut
ta
king
di
sta
nce
betwee
n
nodes
int
o
co
ns
ide
rati
on.
Th
is
research
c
onsidere
d
the
pro
blem
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
20
88
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
8
, N
o.
6
,
Dece
m
ber
201
8
:
437
4
-
4381
4380
us
in
g
a
Mo
dif
ie
d
Bi
na
ry
Pa
rtic
le
Sw
a
rm
Op
ti
m
iz
ation
(
MB
PSO
)
a
ppr
oach
by
up
dating
t
he
optim
i
zat
io
n
functi
on
to
ta
ke
i
nto
c
on
si
derat
ion
the
distances
betwee
n
the
di
ff
e
ren
t
nodes.
T
he
resu
lt
s
wer
e
posit
iv
e
a
nd
this
ap
proac
h
showe
d
no
ti
ceable
im
pr
ove
m
ent
com
par
ed
to
t
he
be
nch
m
ark
.
The
PD
R
s
howe
d
a
n
appr
ox
im
at
e
increase
of
17.
83%
w
her
ea
s
th
e
E2E
delay
saw
a
n
ap
pro
xim
at
e
decr
ease
of
8.
33%
,
an
d
fi
nally
the th
rou
ghput
increase
d by 5.
79% fr
om
the b
enc
hm
ark
stu
dy.
ACKN
OWLE
DGE
MENT
This
w
ork
is
su
pp
ort
ed
by
Mi
nistry
of
Ed
ucati
on
of
Ma
la
ysi
a,
Gr
a
nt
no
:
FRGS/1/
2015/
ICT0
4/UKM
/0
2/
3
.
REFERE
NCE
S
[1]
Am
ee
n,
S.Q.
a
nd
Muni
y
and
i,
R.
C.
,
2017
.
Im
prove
m
ent
at
Network
Planning
using
Heuri
stic
Algorit
hm
to
Minim
iz
e
Cost
of
Distanc
e
betw
ee
n
Nodes
in
W
ire
le
ss
Mesh
Networks.
Inte
r
nat
ion
al
Journal
of
El
ec
tr
ical
a
nd
Com
pute
r
Engi
n
ee
ring
(IJECE)
,
7(1),
pp
.
309
-
315
.
[2]
Toh,
Ch
ai K .
20
01.
Ad ho
c
m
obi
le
wir
el
es
s n
et
w
orks:
protoc
o
ls and s
y
st
ems
,
Pear
son E
ducation
.
[3]
F.L
.
Kha
le
e
l,
N.
S.
As
haa
ri,
T.
S
.
Meria
m
,
T
.
W
ook,
and
A
.
Ism
ai
l,
"P
rogra
m
m
ing
Le
arn
ing
Requ
ir
ements
Based
on
Multi
Perspe
ct
iv
es"
,
Int
ern
ationa
l
Journal
of Elec
tri
c
al
and
Com
p
ute
r Engineering
,
vol
.
7
,
pp
.
1
-
8
.
,
2017.
[4]
F.
L.
Kha
leel,
T
.
S.M.T
.
W
ook,
N.S.
As
haa
ri
,
a
nd
A.
Ism
ai
l
,
"
Gam
ifi
ca
t
ion
E
l
ements
for
L
ea
r
ning
Applicat
io
ns
"
,
Inte
rna
ti
ona
l
Journal
on
Advanc
e
d
Scie
nce,
Engi
n
ee
ring
and
I
nfor
m
at
ion
Te
chno
lo
g
y
,
in
pre
ss
,
vol.
6,
pp.
868
-
874,
2016b.
[5]
F.L
.
Kha
leel,
N.
S.
As
haa
ri
,
T
.
S.
Meria
m
,
T.
W
ook,
and
A.
Ism
ai
l
,
"Th
e
Archi
t
ec
tur
e
of
D
y
n
a
m
ic
Gam
ifi
catio
n
El
ements
B
ase
d
Le
arn
ing
Cont
en
t"
,
Journa
l
of
Co
nver
gence
Infor
m
at
ion
T
ec
hnolo
g
y
,
vol
.
11
,
pp
.
1
64
-
177,
2016
a.
[6]
F.L
.
Kha
leel,
N.
S.
As
haa
ri
,
T
.
S.
Meria
m
,
T
.
W
o
ok,
and
A.
Ism
a
il
,
"U
ser
-
Enj
o
y
a
ble
Le
arn
ing
En
vironment
Base
d
on
Gam
ifi
cation
El
ements"
,
in
I
nte
rna
ti
ona
l
Con
fer
ence
on
Com
pute
r,
Com
m
uni
ca
t
ion,
and
Con
trol
T
ec
hnolo
g
y
(I4CT
2015)
,
,
Kuchin
g,
Sara
wak
,
Mal
a
y
s
ia,
2015
b,
p
.
221
.
[7]
F.L
.
Khaleel
,
N.S.
As
haa
ri,
T
.
S.
Meria
m
,
T.
W
ook,
and
A.
Ism
ai
l,
"The
stud
y
of
gamificat
ion
app
li
c
at
io
n
arc
hi
te
c
ture
for
progra
m
m
ing
la
nguage
cour
se
"
,
in
Proce
edi
ngs
o
f
the
9th
Inte
rn
a
ti
onal
Conf
ere
n
c
e
on
Ubiquit
ous
Inform
at
ion
Ma
nage
m
ent
and
C
om
m
unic
at
ion, 2
015a,
p
.
17
.
[8]
F.L
.
Kh
al
e
el,
"R
ec
rui
tment and
J
ob
Sear
ch
Appli
ca
t
ion"
,
Univ
ersi
ti
Ut
ara Mal
a
y
si
a,
2011
.
[9]
F.L
.
Khal
eel,
N.
S.
As
haa
ri,
T
.
S.
Meria
m
,
T
.
W
ook,
and
A.
Ism
ai
l
,
"G
amific
at
ion
-
Based
Learni
ng
Fram
e
work
for
a
Program
m
ing
La
nguag
e
"
,
in
Inte
rna
ti
ona
l
C
onfe
ren
c
e
on
El
e
ct
ri
ca
l
Enginee
ring
and
In
form
at
ic
s
(ICE
EI
2017),
La
ngk
awi
,
Ked
ah, Ma
lay
s
ia
,
2018,
In
Pres
s.
[10]
F.L
.
Khaleel
,
N.
S.
As
haa
ri,
T.
S.
Meria
m
,
T
.
W
ook,
and
A.
Ismai
l
,
"M
et
hodolo
g
y
for
Deve
lop
i
ng
Gam
ifi
ca
ti
on
-
Based
Learni
ng
Program
m
ing
L
angua
ge
Fram
e
work"
,
in
Inte
rn
at
ion
al
Confer
en
ce
on
Elec
tri
c
al
Engi
ne
eri
ng
and
Inform
at
ic
s (IC
E
EI
2017),
La
ngk
a
wi,
Ked
ah, Ma
lay
sia
,
2018
,
In
Pr
ess.
[11]
Cheng,
Ho
Ti
ng
and
W
ei
hu
a
Zh
uang
.
2009b
.
Q
oS
-
drive
n
node
coope
ra
ti
ve
reso
urc
e
al
lo
cation
f
or
wire
l
ess
m
esh
net
works
with
s
e
rvic
e
diff
ere
nt
ia
t
ion.
Glob
al T
e
lecom
m
unic
at
ions
Confer
en
ce,
GL
OBECOM
2009,
IEEE.
[12]
Gupta,
Bhupend
ra
Kum
ar,
Patna
ik,
Srikan
ta,
Mall
i
ck,
Mana
s
Kum
ar,
&
Na
y
ak
,
Ajit
Kum
ar.
(2017).
D
y
namic
routi
ng
al
gori
th
m
in
wire
l
ess m
esh
net
work.
Inte
r
nat
ion
al
Journa
l of Grid and
Uti
lit
y
Com
puti
ng,
8
(1),
53
-
60
.
[13]
Ak
y
il
d
iz,
Ian
F,
Xudong
W
ang
and
W
ei
li
n
W
ang
.
2005.
“
W
ire
le
s
s
m
esh
net
works:
a
surve
y
”
.
Co
m
pute
r
net
works
47.
4:
445
-
487
.
[14]
Ze
na
lda
n,
Fera
s
,
Hass
an,
Suha
idi
,
&
Habba
l,
Adib.
(2017).
Vert
ical
Hand
over
in
W
irele
ss
Hete
roge
neo
us
Networks.
Journal
of
T
el
e
communic
a
ti
on,
E
le
c
tr
onic
and
Com
pute
r Engineering
(
JTEC),
9(1
-
2), 8
1
-
85.
[15]
Nle
y
a,
Sind
iso
M.
(2016).
Desi
gn
and
opti
m
isa
ti
on
of
a
low
co
st
Cognit
ive
Me
sh
Netw
ork.
Univer
sit
y
of
C
ape
Town.
[16]
Saini
,
Jat
inde
r
S
ingh,
&
Sohi,
B
al
winder
Singh
.
(2016).
A
Survey
on
Chann
el
As
signm
ent
Te
ch
nique
s
of
Multi
-
Radi
o
Mul
ti
-
ch
a
nnel
W
ireless M
esh
Network. In
dia
n
Journa
l
of
Scie
n
ce a
nd
Tech
nolo
g
y
,
9(42)
.
[17]
Dua,
Am
it
.
(20
16).
Eff
i
ci
e
nt
Data
Diss
emina
ti
on
in
Vehi
cular
Ad
Hoc
Net
works
.
THAP
A
R
UN
IVERSIT
Y,
PA
TIAL
A.
[18]
Malm,
Nicolas.
(
2016).
Ultr
a
-
re
liable Net
work
-
co
ntrol
le
d
D2D
.
[19]
Shaukat
,
Us
m
an,
Ahm
ed,
Ej
a
z
,
Anw
ar,
Z
ahi
d
,
&
Xia
,
Feng
.
(2016).
Cloudl
et
d
epl
o
y
m
ent
i
n
local
wir
eless
net
works
:
Moti
vat
ion
,
ar
chi
t
ect
ure
s,
app
li
c
at
io
ns,
and
open
cha
l
le
nges.
Journal
of
Ne
twork
and
Com
pute
r
Applic
a
ti
ons,
62
,
18
-
40
.
[20]
Bel
lalt
a,
Boris
,
Bononi,
Lu
ci
an
o,
Bruno,
R
aff
a
el
e
,
&
Kass
le
r
,
Andrea
s.
(2016)
.
Next
g
ene
ra
tion
IEE
E
802
.
1
1
W
ire
le
ss
Lo
ca
l
Area
Networks:
Curre
nt
st
at
us,
f
uture
d
irecti
ons
and
open
ch
al
l
e
nges.
Com
pute
r
Com
m
unic
at
ions,
75,
1
-
25
.
[21]
Meng,
Tong,
W
u,
Fan,
Yang,
Z
heng,
Chen
,
Guihai
,
&
Vasi
la
ko
s,
Athana
sios
V.
(2016).
Spati
a
l
reu
sabilit
y
-
aw
a
r
e
routi
ng
in
m
ult
i
-
hop
wire
l
ess ne
t
works
.
IEEE
Tr
a
nsac
ti
ons on
Co
m
pute
rs,
65(1),
244
-
255.
[22]
Ibra
him,
Jaww
a
d,
Rehman,
A,
Il
y
as
,
M
Saad
B
in
,
Shehz
ad,
Mohs
in,
&
As
hra
f,
M
ar
y
um
.
(2016).
Optimiza
ti
o
n
an
d
Tra
ffi
c
Mana
g
e
m
ent
in
IEE
E
80
2.
16
Multi
-
hop
Rel
a
y
Sta
ti
ons
u
sing
Gene
tic
and
Priority
Algorithm
s.
Inte
rna
ti
on
al
Journal
of
Com
p
ute
r
Sc
ie
nc
e and
Inform
at
ion
Sec
urity
,
14(7), 599.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
& C
om
p
Eng
IS
S
N:
20
88
-
8708
Wi
rel
ess Mesh
Ne
tw
or
ks B
as
e
d on MBP
SO A
lgo
rit
hm t
o
…
(
Sh
iv
an
Q
as
im
Am
ee
n
)
4381
[23]
Ge,
Xiaohu
,
Tu
,
Song,
Mao,
Guoqiang,
W
an
g,
Cheng
-
Xiang
,
&
Han
,
Ta
o
.
(2016).
5G
ultra
-
dense
c
ellular
net
works
.
I
EE
E
W
ire
le
ss
Com
muni
cations,
23
(1)
,
72
-
79
.
[24]
Ze
ng,
Xi
aopi
ng,
Sun,
Meng,
Jia
n,
Xin,
Du,
Der
ong,
&
Miao,
L
ij
uan
.
(2017).
O
pti
m
al
base
sta
tions
pla
nning
for
Coordina
t
ed
Mu
lt
i
-
Point
s
y
st
em. A
EU
-
Inte
rna
t
ion
al
Journa
l
of
Ele
ct
roni
cs
and
Co
m
m
unic
at
ions,
7
3,
193
-
201
.
[25]
Goudos
,
Sotiri
os
K,
Deru
y
ck,
Ma
rgot,
Pl
et
s
,
Davi
d,
Marte
ns,
Luc,
&
Jos
eph,
W
out.
(2017).
Optimi
za
t
ion
of
Pow
er
Consum
pti
on
in
4G
LT
E
Netw
orks
Us
ing
a
Novel
Bar
ebone
s
Self
-
ad
apt
iv
e
Diffe
ren
t
ia
l
Evo
lut
ion
Algor
it
h
m
.
Te
l
ec
om
m
unic
ation
S
y
st
ems
,
1
-
1
2.
[26]
HA
,
Mahm
oud
Pesara
n,
Hu
y
,
P
hung
Dang,
&
Ramacha
n
dar
amurth
y
,
Vigna
K.
(2016).
A
rev
iew
of
the
opti
m
a
l
al
lo
ca
t
ion
of
distri
bute
d
g
ene
ra
tion:
Objec
ti
v
es,
constra
in
ts,
m
et
hods,
and
al
gori
thms
.
Rene
wabl
e
and
Sus
ta
ina
bl
e
Ene
rg
y
R
evi
ews
.
[27]
Kaur,
R
avne
e
t,
&
Kum
ar,
As
h
wani.
(2016
).
A
n
Approac
h
for
Sele
c
ti
ng
Opt
im
um
Num
ber
of
Base
Sta
ti
ons
a
nd
Optimizi
ng
Site
Loc
at
ions
usin
g
Flower
Polli
nat
ion
Algorit
hm
.
Inte
rna
t
iona
l
Jo
urna
l
of
Com
pu
te
r
Applicati
ons,
133(10),
34
-
39
.
[28]
Sinha,
Kous
hik,
Ghos
h,
Sasthi
C,
&
Sinha,
Bh
aba
ni
P.
(2016)
.
W
ire
le
ss
Networks
and
Mobile
Com
puti
n
g:
CRC
Press
.
[29]
Doppler
,
Klaus,
Reda
na,
Sim
one,
W
ódcz
ak
,
Micha
ł
,
Rost,
Peter,
&
W
ic
hm
an,
Risto.
(2009).
Dy
nami
c
resour
c
e
assignm
ent
and
coope
ra
ti
ve
r
el
a
y
ing
in
c
ellula
r
net
works
:
Concept
and
per
form
a
nce
assess
m
ent
.
EURAS
IP
Journ
al
on
W
ire
le
ss
Co
m
m
unic
at
ions and Net
working
,
2009(1),
475281
.
[30]
Chen,
Chi
‐
Yuan
,
Tseng,
Fan
‐
Hs
un,
Lai,
Chin
‐
F
eng,
&
Chao
,
H
an
‐
Chie
h
.
(2015
).
Network
pla
n
ning
for
m
obil
e
m
ult
i
‐
hop
re
lay
net
works
.
W
irel
ess Com
m
unic
at
ions a
nd
Mobi
le
Com
puti
ng,
15(
7),
1142
-
1154
.
[31]
Murugada
ss
,
Arthi,
&
Pa
chiy
ap
pan,
Arulm
ozhivarm
an.
(2017).
Fuzz
y
Log
ic
B
ase
d
Covera
g
e
a
nd
Cost
Eff
ec
t
iv
e
Plac
ement
o
f
Serving
Nodes
f
or
4G
and
Be
yond
Cel
lu
la
r
N
et
works
.
W
ire
l
e
ss
Comm
unic
at
i
ons
and
Mobile
Com
puti
ng,
201
7.
[32]
W
ang,
Yupeng,
Su,
Xin,
Choi
,
Dongm
in,
&
Choi,
Chang
.
(201
6).
Coordinated
Schedul
ing
Alg
orit
hm
for
S
y
ste
m
Util
ity
Maximi
z
at
ion
W
it
h
Het
er
og
ene
ous
QoS
R
equi
rement
s
in
W
ire
le
ss
Rel
a
y
Networks.
IEE
E
Acc
ess,
4,
8351
-
8361.
[33]
Vasishta,
Anuj
,
Gza
ra
,
Fatma,
Ho,
Pin
-
Han,
&
Li
n,
B
in.
(201
6).
Optimal
loc
a
ti
on
pl
anni
ng
of
rel
a
y
-
b
ase
d
ne
x
t
gene
ra
ti
on
wi
reless
ac
c
ess ne
two
rks.
W
ire
l
ess Ne
tworks,
22(7), 2
159
-
2172.
[34]
Chang,
Jau
-
Yan
g,
&
Chen,
Yun
-
W
ei
.
(2016).
A
rel
a
y
sta
ti
on
depl
o
y
m
ent
sche
m
e
with
a
rota
ti
onal
c
luste
rin
g
al
gorit
hm
for
m
ult
i
-
hop
r
ela
y
n
et
works
.
Paper
pre
sente
d
at
th
e
S
y
stem
Sci
ence
and
Engi
n
ee
r
i
ng
(ICSS
E),
20
16
Inte
rna
ti
ona
l
Co
nfe
ren
c
e
on
.
[35]
Yan,
Y
ang,
Huang,
Ji
anwe
i,
&
W
ang,
Jing.
(2
013).
D
y
n
amic
bar
gai
n
ing
for
r
el
a
y
-
b
ase
d
coop
era
t
ive
spe
ct
rum
sharing.
I
EEE
Jo
urna
l
on
Sel
ec
t
e
d
Area
s
in
Com
m
unic
at
ions,
31(
8),
1480
-
1493
.
[36]
Shim
,
Ky
usung,
Do,
Nhu
Tri
,
&
An,
Beongku
.
(2017).
Perfor
m
anc
e
Anal
y
s
is
of
Phy
s
i
ca
l
Lay
er
Se
cur
ity
o
f
Opportunisti
c
Sc
hedul
ing
in
Mul
t
iuser
Mult
ire
l
a
y
Coopera
t
ive
N
etw
orks.
Sensors
,
17(2),
377
.
[37]
Mata
m
,
Rak
esh
and
Som
ana
th
Tri
pa
th
y
.
2013
.
“
Im
prove
d
he
uristi
cs
for
m
ul
ti
c
ast
ro
uti
ng
i
n
wire
le
ss
m
esh
net
works
”
.
W
ireless ne
tworks 19
.
8:
18
29
-
1837
.
[38]
Li
ng,
Song
,
Ca
o
Jie
and
Yang
Xue
-
jun
.
2010.
Multi
-
pa
th
an
yca
st
rout
ing
b
ase
d
on
an
t
co
lon
y
op
ti
m
iz
a
ti
on
i
n
m
ult
i
-
gatewa
y
W
MN
.
2010
5th
Inte
rn
at
ion
al Co
nfe
ren
c
e
on
Co
m
pute
r
Scie
n
ce
and
Edu
ca
t
ion.
I
CCS
E,
IE
EE.
[39]
Jia,
Jie
,
et
a
l.
.
20
12.
Tra
ff
ic
awa
r
e
r
esourc
e
a
ll
o
cation
for
through
put
opti
m
iz
a
ti
on
in
cogni
ti
v
e
rad
io
wire
le
ss
m
esh
net
works
.
2012
7th
Int
ern
ationa
l
S
y
m
posium
on
W
ire
le
ss
and
Per
vasive
Com
puti
n
g.
ISW
PC
,
IEEE
.
[40]
Xhafa
,
Fatos,
A
dm
ir
Barol
li
and
Makoto
Ta
ki
zaw
a
.
2011.
A
ta
b
u
sea
rch
a
lgori
th
m
for
eff
ic
ie
n
t
n
ode
place
m
ent
i
n
wire
le
ss
m
esh
net
works
.
2011
T
hird
Inte
rn
at
ion
a
l
Confer
en
ce
on
Inte
lligen
t
Netw
orking
and
Col
laborat
iv
e
S
y
st
ems
.
INCoS
,
IE
EE
.
[41]
Ernst,
Jason
B.
a
nd
Tha
bo
Nkw
e
.
2010.
Adapt
ive
m
ixe
d
bia
s
resourc
e
alloc
at
ion
f
or
wire
le
ss
m
esh
net
wor
ks.
2010
Inte
rna
ti
ona
l
Co
nfe
ren
c
e
on
Bro
adba
nd,
W
ire
l
ess Com
puti
ng,
Co
m
m
unic
at
ion
an
d
Applicati
ons
.
BW
CCA,
IEE
E
.
[42]
Khale
d,
Shah
Mos
ta
fa
2012.
Heuri
stic
al
gor
ithm
s
for
wire
le
ss
m
esh
net
work
pla
nning
,
Diss
.
Le
thbr
idge,
Alt
a.
:
Univer
sit
y
of
L
e
thbri
dge
,
Dep
t. o
f
Mathe
m
atics
a
nd
Com
pute
r
Sci
enc
e
,
2012
.
[43]
Paschos,
Georgi
os
S,
Pett
eri
M
anne
rsalo
and
Thomas
Micha
e
l
Bohnert.
2008
.
Cell
ca
p
acit
y
for
ie
e
e
802.
16
cove
rag
e ext
ensi
on.
Consum
er
C
om
m
unic
at
ions
and
Networki
ng
Confer
ence, 200
8.
CCNC
2008
.
5th
IE
EE,
IE
EE.
[44]
Li
n,
Hu
a,
San
th
oshkum
ar
Sam
b
amoorth
y
,
Sande
ep
Shukla
,
Jam
e
s T
horp
and Lam
ine
Mili
.
2011.
Pow
er
s
y
stem a
nd
comm
unic
at
ion
net
work
co
-
sim
ula
ti
on
for
sm
ar
t
grid
applicatio
ns.
Inn
ovat
iv
e
Sm
art
Grid
Te
c
hnologi
es
.
ISG
T
,
2011
IEEE
PES
,
IEEE.
[45]
Mirja
lili
,
Se
y
ed
al
i
,
&
Le
wi
s
,
A
ndre
w.
(2013)
.
S
-
shape
d
ver
sus
V
-
shape
d
tr
ansfe
r
func
ti
ons
fo
r
bina
r
y
par
t
ic
l
e
sw
arm opt
imiza
t
ion.
Sw
arm a
nd
Evol
uti
on
ar
y
Co
m
puta
ti
on,
9,
1
-
14.
[46]
Mirja
lili
,
Se
y
ed
al
i
,
&
Le
wis
,
A
ndre
w.
(2013)
.
S
-
shape
d
ver
sus
V
-
shape
d
tr
ansfe
r
func
ti
ons
fo
r
bina
r
y
par
t
ic
l
e
sw
arm opt
imiza
t
ion.
Sw
arm a
nd
Evol
uti
on
ar
y
Co
m
puta
ti
on,
9,
1
-
14.
[47]
Jun,
Jange
un
an
d
Mihai
l
L
Sic
hit
iu.
2003.
“
Th
e
nom
ina
l
ca
pa
ci
t
y
of
wi
reless
m
esh
net
works
”
.
I
EE
E
wi
reless
comm
unic
at
ions
.
10
.
5:
8
-
14
.
[48]
Cheng,
Ho
Ti
n
g
and
W
ei
hua
Zhua
ng.
2009
a.
“
QoS
-
drive
n
MA
C
-
lay
e
r
reso
urc
e
a
ll
oc
at
ion
for
wire
le
ss
m
esh
net
works
with
non
-
al
tru
isti
c
n
ode
coope
ra
ti
on
and
servic
e
diffe
ren
ti
a
ti
on”
.
IEE
E
Tr
ansa
cti
ons
on
W
ire
le
ss
Com
m
unic
at
ions
.
8.
12:
6089
-
61
03.
Evaluation Warning : The document was created with Spire.PDF for Python.