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
,
Decem
ber
201
8
, p
p.
474
5
~
4754
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
8
i
6
.
pp
474
5
-
47
54
4745
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
Cross L
ayer Sol
ution for
Energ
y and Del
ay Op
timiza
tion
in MAN
ETs
Bhagy
as
hri
R Han
ji
1
, Ra
j
ashre
e She
ttar
2
1
Depa
rtment of
Com
pute
r
Scie
n
ce
Engi
ne
eri
ng
,
Global
Ac
ade
m
y
of
T
ec
hno
log
y
,
India
2
Depa
rtment of
Com
pute
r
Scie
n
ce
Engi
ne
eri
ng
,
R
V Col
l
ege of Engi
ne
eri
ng
,
Ind
ia
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Ja
n
1
, 201
8
Re
vised
Jun
18
, 201
8
Accepte
d
J
ul
28
, 2
01
8
A
novel
m
et
hod
for
pac
k
et
forw
a
rding
in
MA
NE
Ts
has
be
en
pro
posed
in
th
is
pape
r.
A
node
in
the
net
work
a
cts
as
both
host
and
route
r.
Ene
rg
y
uti
lization
of
the
nod
e
inc
r
ea
ses
as
al
l
node
s
in
MA
NET
op
era
t
e
as
sourc
e,
desti
nation,
and
route
r
to
for
ward
pac
ke
ts
to
the
next
hop
ulti
m
at
ely
to
re
ac
h
desti
nation.
Route
r
s
exe
cu
te
a
var
ie
t
y
of
fun
ct
ions
from
simpl
e
pac
k
et
class
ifi
cation
for
forwardi
ng
to
complex
pa
y
lo
a
d
rev
ision.
As
the
num
ber
o
f
ta
sks
and
complexi
t
y
in
cr
ea
ses,
proc
essing
ti
m
e
req
uir
ed
al
so
inc
re
ase
s
result
ing
in
signifi
c
ant
proc
e
ss
ing
del
a
y
in
ro
ute
rs.
Th
e
propo
sed
work
opti
m
i
ze
s
pac
ke
t
hea
der
at
tr
ansport
and
net
work
lay
er
b
y
c
alculat
ing
Unique
Ide
n
ti
fier
using
pai
ring
fun
ct
ion
for
the
f
ields
which
do
not
ch
a
nge
for
a
source
–
desti
nation
pai
r.
Th
is
te
chn
ique
opti
m
izes
the
proc
essing
cost
of
ea
ch
pa
cke
t
he
a
de
r
the
reb
y
conse
rvi
ng
ene
rg
y
and
r
educ
ing
d
ela
y
.
I
t
al
so
sim
pli
f
ie
s
the
ta
sk
o
f
s
y
stem
administ
rat
ion
.
Thi
s
pap
er
el
uc
idates
an
ext
ension
to
ba
sic
AO
DV
protoc
ol
,
al
lowi
ng
routi
ng
of
m
ost
pac
ket
s
without
an
exp
li
cit
hea
d
er,
red
uci
ng
th
e
o
ver
hea
d
of
th
e
pro
toc
ol
whil
e
stil
l
conse
rv
in
g
it
s
basic
prope
rties.
The
proposed
m
ethod
improves
the
n
et
work
p
erf
orm
anc
e
signifi
c
ant
l
y
co
m
par
ed
to
AO
DV
,
MTPR
,
and
S
-
AO
DV
protoc
ol.
Ke
yw
or
d:
AOD
V
Energy m
anage
m
ent
MANET
Ov
e
r
head re
duct
ion
Pairin
g
f
unct
io
n
Rou
te
optim
iz
a
ti
on
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
:
Bhagyas
hr
i R
Hanji,
Dep
a
rtm
ent o
f C
om
pu
te
r
Scie
nce a
nd E
ng
i
ne
erin
g,
Global
A
ca
de
m
y of
Tec
hnol
og
y,
Be
ng
al
uru,
Ka
rn
at
a
ka,
India
.
Em
a
il
:
bh
agya
sh
ri
@g
at
.ac
.in
1.
INTROD
U
CTION
Mob
il
e
A
d
H
oc
NET
work
(
MANET
)
is
a
sel
f
-
go
vernin
g
gro
up
of
sca
tt
ered
m
ob
il
e
us
ers
,
w
he
re
ever
y
node
op
erates
as
s
ourc
e,
de
sti
nation,
and
al
so
pas
se
s
on
pac
kets
f
or
ot
her
no
des,
hen
ce
act
s
as
a
router
al
so
.
T
he
i
dea
of
Mo
bile
A
d
Ho
c
Net
wor
king
is
to
m
ain
ta
in
rob
us
t
a
nd
pro
fici
e
nt
op
e
rati
on
in
m
ob
ile
wireless
netw
orks
by
i
nclu
ding
routin
g
f
unct
ion
al
it
y
into
m
ob
il
e
nodes
[
1]
,
[
2].
MA
NE
T
s
ha
ve
gaine
d
m
uch
at
te
ntion
in
res
earch
ai
m
ed
to
i
m
pr
ove
thei
r
basic
pe
rfo
rm
a
nce
a
nd
pro
vidi
ng
Q
ualit
y
of
Ser
vice
(
QoS).
Q
oS
su
pp
or
t
in
M
AN
E
Ts
re
qu
i
r
es
m
or
e
har
m
on
i
ous
co
op
e
r
at
ion
bet
ween
la
ye
rs,
neces
sit
at
ing
exch
a
ng
e
of
inf
or
m
at
ion
be
tween
la
ye
rs
r
at
her
tha
n
c
on
ven
ti
onal
in
de
pende
nt
la
ye
red
net
wor
k
arc
hitec
ture.
The
r
e
is
current
a
nd
fu
t
ur
e
nee
d
for
dy
nam
ic
Ad
H
oc
netw
orki
ng
te
chnolo
gy.
I
ntern
et
P
r
oto
c
ol
(I
P
)
is
t
he
first
ch
oice
of
t
ran
s
port lay
er pr
oto
c
ol in wire
d
an
d wire
le
ss n
et
w
orks
. Gene
rall
y i
t ha
s b
ee
n ob
s
er
ve
d
that t
he head
er s
iz
e
of
the
pac
ket
is
ab
out
the
sa
m
e
siz
e
of
the
pack
et
or
bigg
er
tha
n
t
he
payl
oad
data.
The
protoc
ol
head
e
rs
a
r
e
si
gn
ific
a
nt
f
or
end
-
to
-
en
d
co
nnect
ions
in
volvin
g
m
ulti
ple
hops
w
her
ea
s
t
he
protoc
ol
he
ader
s
are
insi
gnific
ant
and
do
no
t
ser
ve
any
co
ns
tr
uc
ti
ve
purpose
wh
e
n
ther
e
are
no
interm
ediat
e
no
des
a
nd
c
omm
un
ic
at
ing
nodes
are
c
onnected
directl
y
[
3].
O
pti
m
iz
ing
the
pa
cket
hea
der
siz
e
le
ads
t
o
i
m
pr
ov
em
ent
in
e
ff
ic
ie
nt
res
ource
util
iz
at
ion
[4].
In
Rahul
Desai
et
.
al
.
aut
hors
perform
a
com
par
at
ive
analy
s
is
of
va
rio
us
e
xisti
ng
r
ou
ti
ng
prot
oco
ls
su
c
h
as
DSD
V,
A
ODV,
A
OMD
V,
OL
S
R
an
d
D
SR
[5]
.
An
al
ysi
s
s
hows
that
AOD
V
a
nd
D
SR
ar
e
bette
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.
8
, N
o.
6
,
Dece
m
ber
201
8
:
474
5
-
4754
4746
su
it
able
f
or
hi
gh
t
raffic
net
work
at
the
co
st
of
la
r
ge
delay
.
In
S.
S
y
es
et
.
al
.
aut
hors
pro
po
se
a
m
eth
od
t
o
i
m
pr
ove
the
pe
rfor
m
ance
of
AOD
V
an
d
D
SR
reacti
ve
routing
protoc
ol
in
te
rm
s
of
del
ay
and
over
he
ad
[
6]
.
The
li
nk
resid
ual
li
fe
is
est
im
at
e
d
to
re
duce
li
nk
fail
ur
e
an
d
vel
ocity
of
m
ov
in
g
no
de
is
c
on
si
dered
wh
il
e
choosi
ng
the
ne
xt
node
t
o
forw
a
rd
da
ta
.
T
he
aut
hors
pro
pose
a
m
et
ho
d
to
fi
nd
bette
r
routin
g
path
ha
vi
ng
m
or
e
energy
than
ot
her
rout
es
thr
ough
th
e
analy
sis
of
aver
a
ge
e
nerg
y
and
m
ini
m
u
m
ener
gy
of
paths
exten
ding
the
netw
ork
li
feti
m
e
in
[7
]
.
A
n
al
gorithm
to
eff
ect
ively
m
ana
ge
the
c
ons
umpti
on
of
band
w
idth
a
nd
energy
by appl
yi
ng
m
ob
il
e ag
ents to
car
ry th
e d
at
a is
put f
orwa
rd b
y a
utho
rs
in
[8].
The
m
ob
il
e
agen
t
will
m
ov
e
to
the
neig
hbor
no
de
with
t
he
r
ou
te
re
que
st
pack
et
.
It
w
il
l
loo
k
f
or
node
’s
r
oute
c
ache
to
fi
nd
t
he
path
to
t
he
de
sti
nation
node
.
D
ur
i
ng
this
s
earch
proce
ss
t
he
m
ob
il
e
age
nt
will
op
e
rate
in
the
disco
nn
e
ct
ed
m
od
e
and
hence
will
no
t
use
band
width.
Th
us
the
band
width
a
nd
po
wer
a
re
op
ti
m
iz
ed.
In
Rahul
e
t.
al
.
au
thors
desc
ribe
a
new
op
ti
m
izati
on
te
ch
niqu
e
based
on
rei
nfor
cem
ent
le
arn
i
ng
wh
e
re
r
outi
ng
ta
bles
are
re
pl
aced
by
est
im
at
ion
ta
bles
ca
ll
ed
as
Q
Val
ue
s
[9]
.
Q
val
ue
s
are
base
d
on
li
nk
delay
.
T
he
res
ults
sho
w
that
the
nu
m
ber
of
pack
et
dro
ps
an
d
delay
is
even
lo
w
wh
e
n
the
net
wor
k
siz
e
is
increase
d.
I
n
[10],
aut
hors
pr
ese
nt
a
ne
w
ide
ntifie
r
base
d
m
ulti
ho
p
IP
hea
der
com
pr
essio
n
desig
n.
Fu
rt
her
m
or
e
it
al
so
res
olv
e
s
pro
blem
s
with
pure
IP
ba
sed
ad
ho
c
net
wor
ks
em
erg
ed
du
e
to
IP
a
ddres
s
aut
o
config
ur
at
io
n
serv
ic
e,
dist
rib
uted
nam
ing
a
nd
nam
e
res
olu
ti
on
an
d
at
a
pp
li
cat
io
n
la
ye
r
the
r
ole
of
a
n
IP
address as
an
identifie
r
. An
appr
oach
to
lo
w
er n
et
w
orkin
g ov
e
r
head
a
nd
inv
e
sti
gate the co
nse
que
nce of
la
rg
e
r
pack
et
siz
es is
fo
c
us
e
d
in
[1
1].
The
arti
cl
e
re
views
diff
e
re
nt
m
et
ho
ds
to
reduce
ove
rh
e
ad
by
ha
ving
la
rg
e
f
ram
e
si
ze,
interr
upt
coalesci
ng,
an
d
co
py
avo
i
da
nce
by
pa
ge
re
m
app
in
g,
inte
grat
ed
co
py/c
he
cks
um
and
hardw
a
re
com
pu
t
at
ion
.
High
ba
ndwi
dth
in
the
syst
e
m
can
be
achie
ved
by
re
du
ci
ng
host
ove
rh
ea
ds
by
a
pp
ly
in
g
op
ti
m
iz
a
ti
on
s
above
and
belo
w
the
require
d
la
ye
r,
ide
ntifyi
ng
t
he
net
work
i
nterf
ace
s
upport.
In
te
r
net
En
gi
neer
i
ng
Tas
k
Fo
r
ce
(I
ET
F)
ha
s
de
velo
ped
tw
o
I
Pv6
ap
plica
bl
e
head
er
c
om
pr
ession
sc
hem
e
s
and
their
ap
plica
bili
ty
to
Global
Inform
at
ion
Gri
d
(
GIG)
in
fr
as
tructu
re
is
disc
us
se
d
in
arti
cl
e
[12].
Seve
ral
chall
enges
nec
essary
to
be
t
houg
ht
of
before
ap
pl
yi
ng
hea
de
r
co
m
pr
ession
to
a
par
ti
cular
de
vi
ce
is
al
so
discuss
e
d.
A
no
vel
pack
et
forw
a
r
ding
key ide
ntific
at
ion sc
hem
e in optic
al
p
ac
ket s
witc
hed n
et
work
s
is prese
nte
d
in
[1
3].
Neit
her
head
e
r
m
od
ific
at
ion
nor
la
bel
distrib
ution
prot
oco
l
is
re
qu
i
red,
m
aking
netw
ork
m
anag
em
ent
sim
ple
and
al
so
reduces
t
he
c
om
po
nen
t
c
os
t.
In
H
ooshi
ar
a
uth
ors
put
f
orwa
rd
t
unnelli
ng
ba
sed
route
op
ti
m
iz
a
ti
on
,
w
hich
re
du
ce
s
pack
et
ov
e
r
head
an
d
com
m
un
ic
at
ion
delay
[
14
]
.
By
te
s
con
s
ume
d
t
o
est
ablish m
ob
il
e comm
un
ic
at
i
on
is the k
ey
c
om
par
ison
m
etr
ic
u
sed
. Th
e r
esults sh
ow
s th
at
m
or
e p
ay
loads
are
transm
itted
as
pack
et
s
sent
c
on
ta
in
le
ss
pac
ket
ove
rh
ea
d.
I
n
Fahim
and
Vila
s
auth
ors
ai
m
to
s
peed
up
t
he
data
transm
issi
on
,
r
edu
ci
ng
the
pa
cket
pr
ocessin
g
ti
m
e
by
op
ti
m
iz
ing
the
c
urren
t
IP
pac
ket
structu
re
a
nd
r
edu
ci
ng
head
e
r
siz
e
[15]
.
In
Bow
-
Nan
et
.
al.
auth
or
s
introdu
ce
M
A
NET
I
P
Hea
de
r
Com
pr
essio
n
(MIP
HC)
pro
tocol
that
reduces
20
byte
s
hea
der
to
8
byte
s
[16
]
.
Var
i
ou
s
hea
der
c
om
pr
essio
n
a
nd
dec
om
pr
ession
te
ch
niques
a
r
e
introd
uced
ea
rlie
r,
but
com
pr
essing
a
nd
decom
pr
essing
in
volves
var
io
us
s
te
ps
to
be
pe
rf
or
m
ed
increasi
ng
th
e
processi
ng tim
e an
d
e
nergy c
on
s
um
ption
.
In
M.
Ta
m
il
ar
a
si
et
.
a
l.
a
uthor
s
reco
m
m
end
to
co
ns
e
rv
e
batte
ry
power
by
transm
itti
ng
pa
ckets
wit
h
m
ini
m
u
m
req
uire
d
ene
rg
y
gen
e
rati
ng
le
ss
ov
e
r
head
a
nd
delay
com
par
ed
t
o
sta
nd
ard
DS
R
[
17]
.
Thi
s
p
er
form
ance
is
achieve
d
by
m
od
ify
ing
pac
ket
hea
der
for
m
at
wh
ic
h
exc
lud
es
detai
ls
of
interm
ediat
e
nodes
retai
ning
only
so
urce
a
nd
de
s
ti
nation
a
ddres
s.
Mi
nim
u
m
T
otal
Tra
ns
m
iss
ion
P
ower
R
ou
ti
ng
(MT
PR)
wh
ic
h
con
ce
ntrates
on
en
d
-
to
-
e
nd
e
nergy
eff
ic
ie
nc
y
sel
ect
in
g
the
m
ini
m
u
m
ho
p
path
has
been
addresse
d
in
[
18
]
-
[
19]
w
hile
rem
ai
nin
g
en
er
gy
of
no
de
is
no
t
t
aken
into
acc
ount
f
or
decisi
on
m
aking
.
L
oc
at
ion
a
nd
the
a
ngular
disp
la
cem
ent o
f
nodes wh
il
e sele
ct
ing
the interm
ediat
e n
od
e
s h
as b
ee
n
co
nsi
der
e
d
in [20
]
.
Th
is ens
ures s
ta
ble
path
bet
ween
s
ource
an
d
dest
inati
on
in
tu
rn
reducin
g
the
f
r
equ
e
nt
li
nk
-
bre
aks
an
d
is
devel
op
e
d
over
A
ODV
protoc
ol.
The
m
et
ho
d
cal
le
d
sta
ble
an
d
re
duced
li
nk
br
ea
k
routin
g
protoc
ol
(
S
-
AOD
V)
r
edu
ce
s
the
pr
oc
ess
of
fin
ding the
n
e
w path
oft
enly
.
Sele
ct
ing
m
in
i
m
u
m
ho
p
pa
th,
sen
ding
la
rg
er
pac
ket
siz
e,
us
in
g
head
e
r
com
pr
essi
on
an
d
deco
m
pr
essi
on
are
the
diff
e
r
ent
m
e
tho
ds
use
d
to
reduce
head
e
r
ov
e
rh
e
ad.
C
on
ce
ntrat
ing
on
m
ini
m
um
hop
paths
al
ways
a
tt
e
m
pts
to
sel
ect
nodes
ly
in
g
at
the
ce
nter
and
their
i
nvol
vem
ent
in
f
orwardin
g
data
i
s
al
so
consi
der
a
bly
hi
gh
.
Inv
olv
in
g
la
rg
er
or
var
ia
ble
pac
ket
siz
e
increases
the
per
ce
ntage
of
data
sent
pe
r
a
tt
e
m
pt.
In
a
fa
ulty
en
vi
ronm
ent,
retra
ns
m
issi
on
of
l
arg
e
r
packet
s
com
es
with
co
st
an
d
rec
om
pu
ta
ti
on
.
Com
pressi
ng
an
d
dec
om
pr
e
ssing
re
duces
t
he
ov
e
rall
siz
e
to
be
tra
ns
m
itted
but
t
he
overh
ea
d
i
nvolv
e
d
with
c
om
pr
essio
n
and d
ec
om
pr
es
sion st
il
l con
ti
nu
e
s.
The
pr
opos
e
d
m
et
ho
d
giv
es
a
si
m
ple
and
e
ff
ect
ive
s
olu
ti
on
by
cal
culat
in
g
uniq
ue
num
ber
base
d
on
the
fiel
ds
of h
e
ader
w
hich
re
m
ai
ns
uncha
ng
ed
f
or
c
omm
un
ic
at
ing
pai
rs
in
vo
l
ving v
ery
le
ss
com
pu
ta
ti
on.
This
nu
m
ber
is
incorp
or
at
ed
in
first
data
pack
et
and
is
no
te
d
dow
n
at
al
l
interm
ediat
e
no
de
s
that
will
be
us
e
d
thr
oughout
the session. A
ll
the
seco
nd
a
ry
no
des
will
be
ide
ntifie
d
by
this nu
m
ber
a
nd
ac
cordin
gly
for
w
ard
i
ng
decisi
on
is
m
a
de.
Th
e
pro
posed
m
et
ho
d
ga
ins
inf
or
m
at
ion
thr
ough
interl
ay
er
com
m
u
nicat
ion
an
d
pro
vid
e
s
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
Cross
La
yer
Soluti
on fo
r E
ne
rg
y
and Del
ay Op
ti
miz
atio
n
…
(
Bhagy
as
hri
R
Hanji
)
4747
op
ti
m
iz
ed
path
with
resp
ect
two
pa
ram
et
e
rs
nam
el
y
ene
rg
y
an
d
delay
and
is
buil
t
ov
e
r
A
O
DV
r
eact
ive
routin
g pro
toc
ol, nam
ed
as Optim
iz
ed
–
AODV (
O
-
A
ODV
).
2.
PROP
OSE
D
METHO
D
Faste
r
netw
orks
tra
ns
m
i
t
m
or
e
pack
et
s
a
nd
at
the
sam
e
tim
e
transm
it
m
or
e
payl
oad
byte
s
of
sp
eci
fied
pac
ke
t
siz
e
per
uni
t
tim
e,
increasing
both
per
-
pa
cket
an
d
pe
r
-
byte
over
hea
d
on
the
host
.
Huge
portio
n
of
e
nergy
and
ti
m
e
is
consum
ed
during
the
tra
ns
m
i
ssion
or
rec
epti
on
of
pack
et
w
hich
is
us
e
d
on
ly
for
the
data
adm
in
ist
rati
on
pur
pose
i.e.
fo
r
pr
oc
essing
of
hea
de
rs
an
d
trai
le
rs
[2
1].
T
o
prolo
ng
the
ser
vice
tim
e
of
dev
ic
es
ef
fici
ent
us
ag
e
of
e
n
erg
y
is
a
req
ui
sit
e.
A
series
of
pa
ckets
exc
hange
d
betwee
n
the
sam
e
pair
of
so
urce
de
sti
nation
I
P
ad
dr
ess
es,
port
ad
dr
es
ses
and
us
in
g
the
sam
e
transpor
t
la
ye
r
proto
col
can
be
trea
te
d
i
n
the
sim
i
la
r
m
ann
e
r
by
a
in
te
rm
ediat
e
node
for
for
wardin
g
once
the
deci
sio
n
is
m
ade
base
d
on
first
pack
et
[
22]
.
So
instea
d
of
se
nd
i
ng
the
sam
e
rep
eat
ed
data
fiel
ds
,
th
os
e
fi
el
ds
can
be
re
pl
aced
by
an
ide
ntifie
r
fiel
d.
T
he p
rop
os
e
d
m
et
ho
d,
O
-
AOD
V wor
ks
i
n
tw
o
ste
ps:
Firstl
y
when
th
e
data
is
recei
ve
d
from
hig
he
r
la
ye
r,
pr
im
ary
pac
ket
is
c
reat
ed
with
al
l
t
he
necessa
ry
cal
culat
ion
s
do
ne
a
nd
sent
t
o
lowe
r
la
ye
r
f
or
furthe
r
proces
sing
a
nd
tra
nsm
issi
on
.
Nex
t,
al
l
data
receive
d
will
fit
into
seco
nd
ary
pac
kets
wi
th
com
pact
he
ader
ad
de
d
a
nd
se
nt
to
lo
we
r
la
ye
r
for
tra
nsm
issi
on
as
show
n
i
n
Figure
1.
T
he
head
e
r
of
pr
i
m
ary
pack
et
is
gr
eat
er
t
han
s
econda
ry
pac
ke
t
head
e
r.
E
ve
ry
interm
ediate
node
wh
ic
h
first
rec
ei
ves
the
pr
im
ary
pack
et
sto
r
es
al
l
the
details
in
it
s
ro
utin
g
ta
ble
to
further
ide
ntify
the
flow.
The
sec
onda
ry p
acket
receive
d
ne
xt w
il
l be
check
e
d wit
h
the uniq
ue
ide
nt
ifie
r
if m
a
tc
hed
it
w
il
l be f
or
warde
d
furthe
r
to
ne
xt
node.
T
his
m
eth
od
le
ads
t
o
le
ss
processi
ng
t
i
m
e
and
the
he
ader
siz
e
is
al
so
com
pact
reducin
g
the ove
rh
ea
d d
at
a
wh
ic
h
is ca
rr
ie
d wit
h eve
r
y packet
ulti
m
at
el
y redu
ci
ng
the en
e
r
gy con
su
m
ption
.
Figure
1
.
Pr
im
ary an
d Seco
ndary
Packet tr
a
ns
m
issi
on
Pr
im
ary
and
seconda
ry
pac
ke
t
fo
rm
at
s
are
detai
le
d
in
F
i
gure
2.
Fe
w
fi
el
ds
rem
ai
n
con
sta
nt
f
o
r
a
com
m
un
ic
at
ion
ses
sio
n
bet
ween
tw
o
node
s.
F
or
e
xam
ple,
the
source
and
destinat
io
n
port
i
n
UDP
hea
der
rem
ai
ns
con
st
ant.
T
hese
tw
o
fiel
ds
a
re
t
aken
as
in
pu
t
an
d
uniq
ue
i
den
ti
fier
is
ca
lc
ulate
d.
Sourc
e
a
nd
destinat
io
n
IP
Addresses
rem
ai
n
sam
e
in
IP
head
e
r
f
or
a
c
omm
un
ic
at
ing
pair.
T
hese
tw
o
fiel
ds
a
re
ta
ke
n
as
input
an
d
U
niq
ue
A
ddress
Pair
I
de
ntific
at
ion
Nu
m
ber
i
s
cal
culat
ed
a
nd
sto
red.
Versi
on,
I
nter
net
Head
e
r
Len
gth
,
Ty
pe
of
ser
vice,
Id
e
ntific
at
ion,
Flags,
Tim
e
to
Live
,
P
r
oto
c
ol
rem
ai
ns
uncha
nged
.
Eve
ry
c
omm
un
ic
at
ion
sta
rts
with
pri
m
ary
pack
et
wh
ic
h
c
on
ta
i
ns
al
l
the
fiel
ds
al
ong
with
un
i
qu
e
ide
ntifie
rs.
Eve
ry
interm
ediat
e
n
od
e form
ing
the
path
sto
res
th
e
detai
ls
in
ro
ut
ing
ta
ble
w
hich
will
be
furth
er
us
e
d
to
ide
nt
ify
al
l
seco
nd
a
ry
pac
kets.
Sec
onda
r
y
pack
et
s
f
ollow
t
he
pri
m
ar
y
pack
et
co
nta
ining
c
om
pacted
pac
ket
hea
der
a
nd
increase
d
payl
oad
m
ai
ntaining
sam
e
fr
agm
e
nt
siz
e
at
the
l
ow
e
r
la
ye
rs.
M
bit
wh
en
1
is
us
ed
to
in
dicat
e
the
first/m
idd
le
fr
a
gm
ents
and
la
s
t
fr
agm
ent
when
set
to
0.
Fi
rst
fiel
d
P/S
bi
t
is
us
ed
to
i
ndic
at
e
Pr
i
m
ary
Packet
(P
/S=1
)
a
nd Se
conda
ry Pac
ket (
P/S=
0).
Packet
hea
der
structu
re
is
m
od
ifie
d
as
sh
ow
n
in
F
ig
ure
3,
s
o
as
to
w
ork
acco
r
ding
to
the
requirem
ents.
The
am
ou
nt
of
data
sto
red
in
routin
g
ta
bl
e
al
so
dif
fer
s
to
suppo
rt
the
requirem
ents.
Th
e
incom
ing
pac
ket
is
fir
st
ch
ecked
f
or
pri
m
ary
or
sec
onda
ry
with
th
e
first
fiel
d
(
P/S)
a
nd
the
pack
et
i
s
processe
d
acc
ordin
gly.
F
or
ever
y
pr
im
ary
pack
et
receiv
ed
m
os
t
of
th
e
detai
ls
al
ong
with
i
de
ntific
at
io
n
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.
8
, N
o.
6
,
Dece
m
ber
201
8
:
474
5
-
4754
4748
nu
m
ber
for
th
e
sessio
n
a
re
s
tore
d
in
e
ach
t
up
l
e
a
nd
the
s
econda
ry
inc
om
ing
pac
kets
are
only
ch
eck
ed
f
or
m
at
ching
i
den
t
ifie
r
an
d f
orwa
rd
e
d
acc
ordin
gl
y.
Figure
2. Pr
im
ary an
d
sec
onda
ry Pac
ket H
ea
der Fiel
ds
cl
ass
rt
_e
ntry {
ns
a
ddr_
t
rt_
ds
t
;
ns
a
ddr_
t
r
t
_s
r
c;
u_
i
nt32_
t
r
t
_uid;
u_
i
nt16_
t
rt
_pid;
u_
i
nt8_t
rt_
ps
;
…………
….
}
recv
-
pkt
(P
ac
ke
t *p)
{
Head
e
r
_LS* r
h
=
H
ea
de
r_
L
S :: acc
ess(
p)
;
Head
e
r
_I
P
* i
h
=
H
ea
de
r_IP
::
access(
p)
;
Head
e
r
_P
P
* pp =
Hea
der_P
P
::
access(
p)
;
Head
e
r
_S
P
* pp =
Hea
der_S
P
::
access(
p)
;
…………
………
…………
………
………
If(pp
_p
s
==
1)
Process
the
r
e
cei
ved
pac
ket
as prim
ary pack
et
Else
Process
the
r
e
cei
ved
pac
ket
as seco
ndary
pa
cket
…………
………
…………
………
……….
}
Figure
3. Mo
dified pac
ket h
ea
der str
uctur
e
a
nd it
s pro
ce
ssing
The
m
at
he
m
atical
cal
culat
ion
(t
heoreti
cal
)
for
sta
nd
a
r
d
protoc
ol
head
e
r
ove
rh
ea
d
is
as
sho
wn
i
n
Table
1.
I
n
I
P
v4
ve
rsion,
fiel
ds
li
ke
ve
rsion,
In
te
r
net
Hea
der
Le
ngth,
T
ype
of
Ser
vice
,
I
den
ti
ficat
io
n,
flags
,
protoc
ols
rem
a
ins
uncha
nged
.
The
value
s
de
pict
th
at
fo
r
th
e
pr
im
ary
pack
et
the
ov
er
hea
d
is
increase
d
wh
il
e
for
seco
nd
a
ry
pack
et
s
the
ov
erh
ea
d
is
decr
e
ased.
At
the
en
d
the
ov
e
rall
cal
culat
ion
sho
ws
that
the
pro
po
s
ed
m
et
ho
d
pe
rform
s
bette
r.
For
exam
ple
if
sy
stem
transm
it
s
50,
10
0
a
nd
200
IPv
4/UDP
pac
kets
i
n
a
s
ession
wh
ic
h wil
l have o
ne prim
ary Packet a
nd 49,
99,19
9
sec
onda
ry p
ac
kets
res
pecti
vely
.
Table
1
. O
verh
ead calc
ulati
on for
sta
ndar
d protoc
ol h
ea
de
r wit
hout
op
ti
on
s & pa
dd
i
ng
Proto
co
l
Header
Total Pack
et
Header Si
ze
Pack
et
Heade
r
Siz
e
Ov
erhead
Ov
erall
ef
f
icien
cy
calculatio
n
Pri
m
a
r
y
Seco
n
d
ary
Pri
m
a
r
y
Seco
n
d
ary
Header Ove
rhead
Av
erage
IPv4
/TCP
4
0
bytes
6
7
bytes
2
7
bytes
6
7
.5%
(
↑
)
6
7
.5%
(
↓
)
1
0
pack
ets
7
7
.5%
65%
-
7
0
% (
↑
)
2
5
pack
ets
7
1
.5%
5
0
pack
ets
6
9
.5%
1
0
0
pack
ets
6
8
.5%
2
0
0
pack
ets
68%
IPv4
/UDP
2
8
bytes
4
0
bytes
1
2
bytes
4
2
.8 % (↑)
4
2
.8%
(
↓
)
1
0
pack
ets
5
2
.85
%
35%
-
4
0
% (
↑
)
2
5
pack
ets
4
6
.85
%
5
0
pack
ets
4
4
.85
%
1
0
0
pack
ets
4
3
.85
%
2
0
0
pack
ets
4
3
.35
%
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
Cross
La
yer
Soluti
on fo
r E
ne
rg
y
and Del
ay Op
ti
miz
atio
n
…
(
Bhagy
as
hri
R
Hanji
)
4749
Pairin
g
f
un
ct
io
n
[
23
]
is
a
re
ve
rsible
pr
ocess
to
uniq
uely
encode
tw
o
non
-
ne
gative
nu
m
ber
s
i
nto
a
sing
le
uniq
ue
nu
m
ber
.
C
on
si
der
tw
o
po
sit
iv
e
integers
a
a
nd
b,
PairF
unc
[
a,
b]
yi
el
ds
a
sing
le
posit
ive
intege
r
c
un
iq
uely
ass
ociat
ed
with
t
he
input
pair
.
Inver
se
functi
on
UnPairF
un
c[c]
ou
t
pu
ts
th
e
po
sit
ive
inte
ger
pair
[a,
b]
relat
ed
to
gen
e
rati
on
of
c
[
24
]
,
[25].
Pairin
g
f
unct
ion
a
re
bi
je
ct
ion
s
an
d
is
i
m
po
rtant
in
log
ic
,
com
pu
ta
ti
on
and
m
at
he
m
at
ics
on
the
w
hole
[2
6].
Ba
sic
pr
ope
rtie
s
of
a
pairing
f
un
ct
ion
s
a
re
I
nje
ct
ion
,
Ind
uction
a
nd
on
t
o
as
gi
ven
i
n
eq
uatio
n
(
1)
,
(2
)
a
nd
(
3)
re
sp
ect
ively
.
I
nject
ive
functi
on
m
aps
disti
nct
input
argum
ents
to
disti
nct
ou
tp
ut
va
lues.
T
wo
p
ai
rs
ge
ner
at
in
g
t
he
sam
e
identifie
r
is
po
ssi
ble
on
ly
w
he
n
both
pai
r
s
are ide
ntica
l. Onto
o
r
s
urjecti
ve
fun
ct
io
n
m
aps
eve
ry possi
ble v
al
ue
to a
t l
east
one arg
um
e
nt.
2
2
1
1
2
,
1
2
1
,
b
a
b
a
b
b
a
a
(
1)
b
a
b
b
a
a
,
,
(
2)
c
b
a
b
a
,
0
(
3)
Eq
uation
(
4)
s
ta
te
s
that
the
f
un
ct
io
n
ge
ner
a
te
s
a
uniq
ue
num
ber
as
ou
t
put
f
or
any
gi
ve
n
pair
of
integer
num
ber
s
(
Inu
m
1
,
I
nu
m
2
)
as
i
nput.
T
he
outp
ut
gen
e
rated
is
e
qu
al
on
ly
if
tw
o
pairs
of
in
put
are
equ
al
.
Eq
uation
(
5)
st
at
es
that
giv
e
n
a
nu
m
ber
K
a
s
input,
the
n
th
e
two
i
ntegers
(
I
num
1
,
Inu
m
2
)
can
be
ge
ner
a
te
d
as
ou
t
pu
t.
T
he pairi
ng fu
nction s
at
isfie
s equati
on
(4)
a
nd (5).
4
3
2
1
4
3
2
1
,
,
,
,
,
I
n
u
m
I
n
u
m
I
n
u
m
I
n
u
m
t
h
e
n
I
n
u
m
I
n
u
m
I
n
u
m
I
n
u
m
(4)
K
I
n
u
m
I
n
u
m
t
h
a
t
s
u
c
h
NX
N
I
n
u
m
I
n
u
m
p
a
i
r
a
is
t
h
e
r
e
N
f
o
r
e
a
c
h
K
1
1
2
2
,
,
(5)
The
Ca
nt
or
Pa
iring
f
unct
io
n
is
a
pr
im
i
ti
ve
recu
rsi
ve
pairi
ng
f
un
ct
io
n
de
fi
ned
by
equat
io
n
(
6)
an
d
(
7)
be
low.
Both
eq
uatio
ns
descr
ibe
th
e
f
un
ct
io
n
use
d
t
o
pro
duce
U
niq
ue
s
ource
de
s
ti
nation
Pair
I
de
ntifie
r
(
UP
I
)
ta
king
Inum
1,
Inu
m
2
a
s in
pu
t
values
and
N
as t
he
se
t of val
ues.
2
1
,
:
:
I
n
u
m
I
n
u
m
N
N
X
N
(6)
2
2
1
2
1
2
1
1
2
1
:
,
I
nu
m
I
nu
m
I
nu
m
I
nu
m
I
nu
m
I
nu
m
I
nu
m
(
7)
The
li
m
i
ta
ti
on
of
Ca
nt
or
pai
rin
g
f
unct
ion
is
that
the
ra
nge
of
e
ncode
d
resu
lt
do
e
sn
'
t
al
ways
sta
y
within
the
li
m
i
ts
of
2N
bit
int
eger.
If
the
i
nputs
are
tw
o
N
-
bit
integers
,
th
en
there
a
re
2
N
*
(
2
N
-
1)
com
bin
at
io
n
of
possible
in
puts. W
it
h
Pi
ge
on
Hole
Pr
i
nci
ple
we
nee
d
a
n
ou
t
pu
t of
siz
e at
le
ast
2
N
*
(
2
N
-
1)
w
hich
is
e
qu
al
t
o
2
2N
-
2
N
in
ot
her
wo
r
ds
a
siz
e
equ
al
to
2N
bi
t
nu
m
ber
.
Ma
lth
ew
Sz
udzi
k
m
et
ho
d
us
e
d
in
[18]
ov
e
rc
om
es
the
above
li
m
i
ta
ti
o
n
a
nd
is
de
fine
d
by
e
quat
ion
(8)
be
lo
w.
Inu
m
1
an
d
I
num
2
are
tw
o
key
in
pu
t
i
ntegers
ba
sed
on
wh
ic
h
a
uniq
ue
num
ber
is cal
c
ulate
d
a
nd r
et
urne
d.
0
,
:
2
?
2
1
2
2
1
1
1
1
2
1
I
n
u
m
I
n
u
m
w
h
e
r
e
I
n
u
m
I
n
u
m
I
n
u
m
I
n
u
m
I
n
u
m
I
n
u
m
I
n
u
m
I
n
u
m
I
n
u
m
(
8)
Figure
3,
il
lus
trat
e
diff
e
re
nt
un
i
qu
e
num
ber
key
gen
e
rate
d
re
pr
e
sente
d
by
dot
on
the
gr
a
ph
f
or
arou
nd
500
pai
r
of
num
ber
s
t
aken
a
s
ke
y
1
a
nd
key2
.
T
he
nu
m
ber
s
incl
uded
i
n
pa
irs
w
hich
a
re
te
ste
d
are
as
sm
a
ll
as
two
dig
it
num
ber
ti
ll
twel
ve
dig
it
nu
m
ber
s.
At
the
oth
e
r
e
nd
w
he
n
giv
e
n
a
key
value,
the
n
ke
y1
an
d
key2
m
ay
also
be gene
rated
e
asi
ly
.
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.
8
, N
o.
6
,
Dece
m
ber
201
8
:
474
5
-
4754
4750
Figure
3. U
nique
Id
e
ntifie
r ge
ner
at
e
d
f
or
500 dif
fer
e
nt s
ourc
e
-
destinat
i
on pai
rs
3.
RESU
LT
S
A
ND
DI
SCUS
S
ION
The
e
xperim
ent
is
carrie
d
out
in
NS
-
2.3
5[2
7]
.
Si
m
ulati
on
a
rea
ta
ke
n
is
15
00
m
x
1500m
consi
sti
ng
of
50
m
ob
il
e
node
s.
All
the
nodes
a
re
i
niti
al
iz
ed
to
100J
and
no
des
a
re
arb
it
ra
rily
po
si
ti
on
e
d
with
Ra
ndom
Wayp
oin
t
M
obil
ity
Mod
el
wit
h
m
ob
il
it
y
in
the
ra
nge o
f
0m/
s
to
5m
/s.
250m
is
ta
ken
as
c
omm
un
ic
at
ion
rang
e
and
data
rate
as
2Mbps.
A
O
DV
is
us
e
d
as
fun
dam
ental
ro
uting
protoc
ol
and
IEEE
80
2.11
as
m
edium
acce
ss
con
t
ro
l
protoc
ol.
T
he
m
et
ho
d
disc
u
sse
d
is
m
easur
e
d
in
t
erm
s
of
pac
ke
t
ov
e
rh
ea
d,
pa
cket
proce
ssin
g
ti
m
e,
energy
sa
ve
d
duri
ng
this
proc
ess
and
pe
rce
ntage
of
e
xtra
pa
yl
oad
sent.
Th
e
m
et
ho
d
is
com
par
ed
with
A
ODV
,
S
-
A
O
DV
an
d
pro
ves
po
sit
iv
el
y
good
in
te
r
m
s
of
the
th
re
e
m
e
tric
s
packet
ov
e
r
head,
pa
cket
pr
ocessi
ng
tim
e
and
e
ne
rg
y
c
onsu
m
ption
c
on
sidere
d
f
or
e
va
luati
on
.
Pack
et
ov
e
r
head
is
de
fine
d
as
the
pro
portio
n
of
num
ber
of
byte
s
in
hea
der
to
payl
oa
d
le
ng
th
.
Pac
ket
processi
ng
ti
m
e
is
m
easur
ed
as
the
tim
e
the
node
e
nters
at
input
interface
of
a
nod
e
t
o
t
he
ti
m
e
wh
e
n
t
he
pa
cket
is
put
ou
t
at
the
ou
t
put
interface
.
T
ota
l
energy
c
onsum
ed
is
cal
culat
ed
as t
he diffe
re
nce
be
tween
rem
ai
ni
ng en
e
r
gy and
init
ia
l energ
y.
The
grap
hs
in
Figure
4
a
nd
5
sh
ows
that
the
nu
m
ber
of
pa
yl
oad
transm
it
t
ed
are
m
or
e
by
3
%
to
5%
in
A
ODV
with
IP
v4/U
DP
a
nd
IPv
4/TCP.
G
raph
re
su
lt
is
s
how
n
f
or
50
pa
ckets
to
200
pack
et
s
al
ong
X
-
a
xis
and
num
ber
of
byte
s
of
payl
oad
sent
al
ong
Y
-
a
xis.
Payl
oad
car
ried
in
pri
m
ary
packet
will
be
1460byte
s
(15
18
-
40
-
18)
and
sec
onda
ry
pack
et
will
be
1488
byte
s
(
1518
-
12
-
18)
f
or
I
Pv4
pack
et
s.
Payl
oad
ca
r
ried
in
pr
im
ary
pack
et
will
be
1433
byte
s
(
1518
-
67
-
18)
an
d
s
eco
ndary
pac
ket
wi
ll
con
ta
in
1473
byte
s
(
1518
-
27
-
18)
for IPv4/TC
P
.
Figure
4. Payl
oa
d
Se
nt in
AO
DV an
d O
-
A
O
DV w
it
h IP
v4/
UDP
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
Cross
La
yer
Soluti
on fo
r E
ne
rg
y
and Del
ay Op
ti
miz
atio
n
…
(
Bhagy
as
hri
R
Hanji
)
4751
Figure
5. Payl
oa
d
Se
nt in
AO
DV an
d O
-
A
O
DV w
it
h IP
v4/
TCP
Figure
6,
7
a
nd
8
s
hows
t
he
gr
a
phic
al
resu
l
ts
ob
ta
ine
d
f
or
Rou
ti
ng
loa
d,
Energy
Co
nsu
m
pt
ion
an
d
En
d
to
E
nd
de
la
y
of
the
pr
opos
e
d
m
et
ho
d,
O
-
A
ODV
c
om
par
ed
to
A
ODV,
MT
PR,
an
d
S
-
A
O
D
V.
T
he
pr
im
ary
pack
et
head
e
r
f
or
IP
/
UD
P
is
increas
ed
to
40
byte
s
wh
ic
h
is
hi
gh
e
r
the
n
re
gu
la
r
pack
et
header
of
28
byte
s,
at
the
sa
m
e
tim
e
the
seco
nd
a
ry
pac
ket
hea
der
is
reduce
d
to
12
byte
s.
As
the
rati
o
of
pr
im
ary
to
seco
nd
a
ry
pac
kets
in
ge
ner
al
is
ta
ke
n
as
1:
(N
-
1),
w
her
e
N
is
the
total
nu
m
be
r
of
pac
kets
se
nt,
(
N
-
1)
is
th
e
nu
m
ber
of
se
c
onda
ry
pac
kets
fo
ll
ow
i
ng
a
pr
im
ary
pack
e
t
the
r
ou
ti
ng
ov
e
r
head
is
re
du
ce
d
reas
on
a
bly
as
dep
ic
te
d i
n Fi
gure
6.
Figure
6. N
orm
al
iz
ed
Rou
ti
ng L
oa
d Vs
Num
ber
o
f Pac
ket
s
Energy
co
nsu
m
pt
ion
is
al
so
reduce
d
ap
pro
pr
ia
te
ly
as
sh
own
in
F
igure
7.
A
ve
r
age
ene
rg
y
consum
ption
a
t
the
e
nd
of
sim
ula
ti
on
is
ca
lc
ulate
d
a
nd
f
ound
t
o
be
im
pro
ved
by
6%
to
8%
le
ss
e
ne
r
gy
consum
ption
than
S
-
A
O
DV
m
et
ho
d.
Fi
gur
e
8
sh
ows
tha
t
end
to
en
d
de
la
y
is
al
so
red
uce
d
as
the
pack
e
t
proces
si
ng
ti
m
e
is
reduce
d
as the
num
ber
of f
ie
lds
to b
e
c
he
cked
is
m
ini
m
a
l.
On
an
a
ve
rage
arou
nd
3
%
t
o
5%
i
m
pr
ovem
ent
i
s
sho
wn
with
r
espect
to
delay
.
T
he
processi
ng
tim
e
is
redu
ced
wh
ic
h
a
dd
s
to
t
he
im
pr
ov
e
m
ent
in ove
rall
d
el
a
y.
A
ODV
MTPR
S-AO
DV
O
-AODV
No
rm
al
iz
ed
Rou
ti
ng
Load
Packets
No
rm
al
iz
ed
Rou
ti
ng
Load
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.
8
, N
o.
6
,
Dece
m
ber
201
8
:
474
5
-
4754
4752
.
Figure
7. Ene
r
gy Co
nsum
ptio
n Vs
Nu
m
ber
of Pac
kets
Figure
8. En
d
t
o
E
nd
Delay
Vs
Num
ber
of P
a
ckets
4.
CONCL
US
I
O
N
The
w
ork
in
this
pa
per
f
oc
usse
s
on
a
no
ve
l
m
e
tho
d
to
l
essen
pac
ket
processi
ng
ti
m
e.
Delay
is
beco
m
ing
m
ore
an
d
m
or
e
no
te
worthy
as
ne
tworks
put
i
nto
pr
act
ic
e
c
omplex
prot
oco
ls
r
unning
on
routers
.
Mi
ni
m
iz
ing
host
ov
e
rh
e
ad
i
n
handlin
g
packet
s
and
t
heir
da
ta
a
chieve
bette
r
pe
rfor
m
ance.
O
ptim
iz
ed
pa
cket
head
e
r
tra
ns
m
it
s
m
or
e
payl
oads
with
s
m
al
le
r
head
er
s
and
re
duce
d
proces
sin
g
tim
e
increasing
the
eff
ect
ive
ness
of
the
existi
ng
netw
ork.
T
he
pr
im
ary
packet
is
sent
with
regular
head
e
r
an
d
fe
w
a
ddit
ion
al
fiel
ds
,
wh
i
c
h
will
be
proces
sed
by
al
l
interm
ediat
e
no
de
s
store
the
ne
cessary
detai
ls
require
d
to
furthe
r
forw
a
r
d
al
l
the
secondary
pac
kets
belo
ngin
g
to
the
sa
m
e
c
omm
un
ic
at
ion
session.
The
pro
posed
wor
k
sh
ows
i
m
pr
oved
pe
rfor
m
ance
with
resp
ect
to
t
he
t
hr
ee
m
et
rics
rou
ti
ng
ov
e
rh
ea
d,
e
nergy
c
on
s
um
ed
an
d
e
nd
t
o
e
nd
delay
.
Th
e
w
or
k
is
car
ried
out
with
IPv4
a
dd
ressin
g
a
nd
ca
n
be
exte
nded
t
o
I
Pv6
a
ddress
ing
.
The
idea
be
hind
O
-AODV
S-AO
DV
MTPR
A
ODV
Ener
gy
Con
su
m
ption
Packe
ts
Ener
gy
Con
su
m
ed
O
-AODV
S-AO
DV
MTPR
A
ODV
Packe
ts
End
-
To
-
En
d
Delay
En
d
-
to
-
End
Delay
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
Cross
La
yer
Soluti
on fo
r E
ne
rg
y
and Del
ay Op
ti
miz
atio
n
…
(
Bhagy
as
hri
R
Hanji
)
4753
cal
culat
ing
U
ni
qu
e
Addr
e
ss
P
ai
r
Id
e
ntific
at
ion
N
um
ber
us
ing
t
he
IP
a
ddr
esses
can
be
m
od
i
fied
to
incl
ud
e
t
he
Sour
ce
IP
A
dd
ress
a
nd
the
Start
tim
e
of
the
session
w
he
n
pr
im
ary
pack
et
is
ge
ner
at
e
d.
This
a
dds
sec
uri
ty
as
the Uniq
ue N
um
ber
g
ene
rate
d wil
l be
valid
for on
ly
on
e
se
ssion com
m
un
ic
at
ion
.
REFERE
NCE
S
[1]
S.
Corson,J.
Mac
ker
,
"M
obile
Ad
Hoc
Networki
ng(MA
NE
T)
:
Ro
uit
ng
protoc
ol
P
erf
orm
anc
e
Iss
ues
and
eva
luatio
n
Considera
ti
ons
,
" Net
work Worki
ng
Group,RFC: 2501,1999.
[2]
Le
i
Chen
,
"P
roto
col
s
for
Supporting
Quali
t
y
of
Servic
e
in
Mobi
le
Ad
Hoc
Net
works
,
"
Ph.d
The
sis,Unive
rsit
y
of
Roche
ster
,
New
York,2006.
[3]
Sass
an
Aham
adi
,
"The
IEEE
802
.
16m
Converge
n
ce
Sub
-
L
a
y
er
,
"
Mobile
W
iMAX
,
2011.
[4]
EFF
NET
AB,
"
An Int
roduc
t
ion to
IP Hea
d
er
Co
m
pre
ss
ion,
" White
Pap
er,
2004
.
[5]
Rahul
Desai
,
B
P
Pati
l,
Davind
e
r
Pal
Sharm
a,
“
Routi
ng
Protoco
ls
for
Mobile
Ad
Hoc
Network:
A
Surve
y
and
Anal
y
sis”
.
I
ndo
n
esia
n
Journa
l
of
El
e
ct
ri
ca
l
Eng
in
ee
ring
and
Com
pute
r
Sc
ie
nc
e.
v
ol.
7
,
(3)
,
2017
,
pp.
795
-
801
.
[6]
S.
S
y
es
Abdul
S
y
ed
,
T.
Senthil
Kum
ara
n,
”E
x
tr
emel
y
Vibra
n
t
Routi
ng
Schem
e
for
Mobil
e
A
dhoc
Network”,
Indone
sian
Jour
nal
of
E
le
c
tri
c
al
Engi
ne
eri
ng
and
Com
pute
r
Sci
e
n
ce
,
vo
l. 9(2), 201
8,
pp.
306
-
310.
[7]
Hasan
Aw
ni
Sh
aki
r,”Enh
anc
em
ent
of
Ene
rg
y
Control
Routi
ng
Protocol
for
Mobile
Ad
Hoc
Network
Based
on
H
y
brid
Part
icle
Sw
arm
Optimiz
at
ion
with
Ant
Colon
y
-
b
ase
d
E
ner
g
y
Con
trol
Routi
ng”
,
Indon
esia
n
Journal
o
f
El
e
ct
ri
ca
l
Eng
in
ee
ring
and
Com
pute
r
Sc
ie
nc
e,
vo
l.
8
,
(2),
2017
,
pp
.
308
-
314.
[8]
Bindi
y
a
Bhatia
,
M.
K.
Soni,
Parul
Tomar,
”E
x
te
n
ded
Bandwidt
h
Optimize
d
and
Ene
rg
y
Eff
i
ci
en
t
Dy
n
amic
Sourc
e
Routi
ng
Protoco
l
in
Mobile
Ad
-
hoc
Networks”,
Inte
rna
t
iona
l
J
ourna
l
of
El
e
ct
r
ic
a
l
and
Com
pu
te
r
Engi
n
ee
ring
(IJECE),
vo
l. 8,
(3),
2018
,
pp
.
14
60
-
1466.
[9]
Rahul
Desai
,
B
P
Pati
l,
Davinder
Pal
Sharm
a.
Le
arn
ing
Based
R
oute
Mana
geme
nt
in
Mobile
Ad
Hoc
Networks.
Indone
sian
Jour
nal
of
E
le
c
tri
c
al
Engi
ne
eri
ng
and
Com
pute
r
Sci
en
ce
.
vol
.
7
,
(3)
,
2
017,
pp
.
718
-
72
3
.
[10]
Shahrukh
Khalid,Atha
r
Mahboo
b,
Choudhr
y
Fah
ad
Azim,Aqeel
Ur
Rehman,
"D
esign
and
Reali
z
at
ion
of
a
Nove
l
Hea
der
Com
pre
ss
ion
Scheme
for
Ad Hoc
Networ
ks,"
ET
RI
Journal,
vo
l. 38, (5), 2
016,
pp.
922
-
933
.
[11]
Jeffrey
S.
Chase
,
Andrew
J.
Gallati
n
,
Kenne
th
G
.
Yocum
,
"End
S
y
stem
Optimiza
tions
for
High
Sp
ee
d
TCP,"
IE
EE
Com
m
unic
at
ions Ma
gaz
in
e,
200
1,
pp.
68
-
74.
[12]
Emre
Ert
eki
n
,
C
hris
Christou,
Booz
Alle
n
Ha
m
il
ton,
"Inte
rn
et
Protocol
Hea
der
Com
pre
ss
io
n,
Robust
Hea
d
er
Com
pre
ss
ion,
and
The
ir
Applicabil
i
t
y
in
the
Globl
Inform
at
ion
Grid,
"
IEE
E
Com
m
unic
at
ions
Maga
zi
n
e,
2004
,
pp.
106
-
116.
[13]
Henri
k
W
essing,
Henri
k
Christ
ians
en,
Ti
n
a
Fjel
d
e,
L
ars
Ditt
m
ann
,
"N
ovel
Scheme
for
Packe
t
For
wardi
ng
W
it
hou
t
Hea
der
Modifica
ti
ons i
n
Optical
Netwoks,"
Journal
of
L
ight
wave
Te
chno
log
y
,
vol
.
20,
(8),
2002
,
pp.
1277
-
1288.
[14]
Hoos
hia
r
Zol
fag
har
nasa
b,
"R
educing
Packe
t
Over
hea
d
in
Mobil
e
IPv6,"
Inte
rna
t
i
onal
Journal
of
Distribut
ed
an
d
Para
llel
S
y
stems
,
vol. 3,
(3)
,
2012,
pp.
1
-
8.
[15]
Fahim
A
Ahm
e
d
Ghane
m
,
Vila
s
M.
Tha
kar
e
,
"O
pti
m
iz
ation
Of
IPv4
Packe
t’s
Hea
der
s,"
Int
ern
a
t
iona
l
Journal
of
C
om
pute
r
Scie
n
ce
Iss
ues,
vo
l. 10
,
(1),
2013
,
pp.
29
4
-
302.
[16]
Bow
-
Nan
Che
ng,
John
Zue
n
a,J
im
W
hee
le
r,
S
cot
t
Moor
e,
Bri
an
Hung,"MA
NET
IP
Hea
d
er
Com
pre
ss
ion,
"
Proce
ed
ings o
f
t
he
IE
EE Mi
l
it
ar
y
Comm
unic
at
io
ns Confe
renc
e
,
2
013,
pp.
494
-
503
.
[17]
M.
Ta
m
ilara
si
,
S.
Chandr
amathi,
T.
G.
Pala
n
ive
l
u,
"O
ver
hea
d
Re
duct
ion
and
En
erg
y
Mana
g
ement
in
DS
R
fo
r
MA
NETs
"
proc
ee
dings o
f
Thir
d
int
ernati
ona
l
Co
nfe
renc
e
on
COMm
unic
ati
on
,
C
OMSWAR
E
,
B
an
gal
ore
,
2008.
[18]
Scott
D,
Toh
C
,
Cobb
H,"
Perf
orm
anc
e
Ev
al
ua
ti
on
of
ba
tt
er
y
l
ife
awa
r
e
rout
in
g
sche
m
es
for
wire
l
ess
ad
hoc
net
works
,
"
Proc
ee
dings o
f
I
EE
E
ICC
,
vol
.
9
,
200
1,
pp.
2824
-
2829
.
[19]
W
.
Su,
S.
J.
Le
e
,
M
Gerl
a,
"M
obil
ity
Prediction
a
nd
Routi
ng
in
A
d
Hoc
W
ire
le
ss
Networks,"
Inte
rna
ti
on
al
Journa
l
of
Network
Man
age
m
ent
,
vol
.
11
,
(1),
2001
,
pp.
3
-
3
0.
[20]
Bhag
y
ashri
R
Hanji
,
R
aj
ashr
ee
Shett
ar,
"S
t
abl
e
Reduc
ed
L
ink
Brea
k
Routi
ng
Te
chni
qu
e
in
Mobile
Ad
Hoc
Network,
"
Proc
e
edi
ng
o
f
th
e
S
econd
Int
ernati
on
al
Conf
ere
nce I
CACDS
,
Dehra
d
un,
2018.
[21]
Mobin
Ifte
khar
u
l,
"Ene
rg
y
Eff
ici
ent
Packet
Size
Optimiza
ti
o
n
for
wire
le
ss
Ad
Ho
c
Networks,"Ph.d
T
hesis,
Quee
n
Marr
y
Univer
sit
y
of
London
,
Uni
te
d
Kingdom
,
20
13.
[22]
"N
et
work Rout
i
ng:
Algori
thms
,
Protocol
s, and
A
rch
itect
ur
es,
"2
nd
Edi
ti
on
,
E
lsevi
er
,
2010.
[23]
Steve
n
Pig
eon,
"
Pairi
ng
Func
ti
on
s,"
From
Mathwo
rld
-
A Wolfram
W
eb
Resourc
e
,
c
rea
t
ed
b
y
v
Eri
c
W
.
W
ei
ss
te
in.
[24]
Matt
hew
Szud
zi
k,
"A
n
Elega
n
t
Pairi
ng
Functi
on
,
"
Proce
e
dings
of
the
Wolf
ram
Sci
enc
e
NKS
20
06
Confe
renc
e
,
200
6.
Avai
la
b
le
f
ro
m
htt
p:/
/s
zudz
ik
.
com/El
eg
ant
Pair
ing.
pdf.
[25]
Dovier
,
Agos
ti
no
,
Enrico
Pontelli
,
and
Gianf
r
an
co
Ross
i,
"S
et
unifi
c
at
ion",
T
heor
y
and
Pra
ct
i
ce
of
Logi
c
Program
m
ing,
2006.
[26]
Kenne
th
W
Re
gan,
"M
ini
m
um
-
Com
ple
xity
Pai
ring
Functi
ons,"
Journal
of
Co
m
pute
r
and
S
y
s
te
m
Scie
nc
es,
vo
l.
45,
2006,
pp
.
285
-
295.
[27]
Kevin
Fal
l,
Anna
n
Vara
dh
an, T
he
NS
Manua
l
,
Th
e
VIN
T
Proj
ec
t
,
2011.
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.
8
, N
o.
6
,
Dece
m
ber
201
8
:
474
5
-
4754
4754
BIOGR
AP
HI
ES OF
A
UTH
ORS
Bhag
y
ashri
R
Hanji
recei
v
ed
he
r
B.
E
and
Mtech
degr
e
e
from
V
isvesvar
a
y
a
T
echnological
Univer
sit
y
,
Bela
gavi
.
She
is
cur
ren
tly
work
ing
towar
ds
her
Ph.
D.
degr
ee
deal
ing
with
improving
the
Quali
t
y
of
Serv
ic
e
in
Mobil
e
Ad
Hoc
net
works
.
S
he
is
cur
ren
t
l
y
working
as
As
sista
nt
profe
ss
or
at Global Ac
a
dem
y
of
T
ec
hno
log
y
,
B
enga
lu
ru.
Dr.
Raj
ashr
ee
S
het
t
ar
is
cur
r
ent
l
y
working
as
Profess
or,
Dept.
o
f
Com
pute
r
Science
,
R
.
V
Coll
ege
of
Enginee
ring
,
B
enga
l
uru.
Her
rese
a
rc
h
work
foc
uses
on
“
Know
le
dge
Discove
r
y
in
Sem
i
-
struct
ur
ed
Dat
a”.
She
h
as
aro
und
45
pu
bli
c
at
ions
in
var
i
ous
Inte
rna
ti
ona
l
Journals
and
Confer
e
nc
es.
She
has
aut
hore
d
a
book
on
“S
eque
nt
ia
l
Patter
n
Mining
from
W
eb
Log
Data
:
Con
ce
pts,
Te
chn
ique
s
an
d
Applic
ations
of
W
eb
Us
age
Mining”
,
pub
l
ished
b
y
LAMBERT
Aca
demic
Publi
shin
g
compan
y
,
Ger
m
an
y
and co
-
authored
book
cha
p
te
rs
Evaluation Warning : The document was created with Spire.PDF for Python.