Indonesian J
ournal of Ele
c
trical Engin
eering and
Computer Sci
e
nce
Vol. 2, No. 2,
May 2016, pp
. 417 ~ 425
DOI: 10.115
9
1
/ijeecs.v2.i2.pp41
7-4
2
5
417
Re
cei
v
ed
Jan
uary 21, 201
6
;
Revi
sed Ap
ril 14, 2016; Accepted Ap
ril 28, 2016
A Power Efficient Trust Based SecureRouting Scheme
for Mobile Ad-Hoc Networks
MV Rath
nam
m
a*
1
, P Chenna Re
dd
y
2
JNT
U
A, Anantapur-AP-In
dia.
T
e
l::+
91 9849
7
778
31
JNT
UCEP, Pulivendula-YSR-
AP-India
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: mvrathnamm
a
@gma
il.com
1
, pcredd
y1
@re
d
iffmail.com
2
A
b
st
r
a
ct
MANET
s
are s
e
lf-org
ani
z
i
ng,
infrastructure
l
e
ss
ad-
hoc
net
w
o
rks w
i
th ma
ny cha
l
l
eng
es
like
lo
w
pow
er, li
mite
d
storage
an
d l
i
m
ite
d
pr
ocess
i
ng d
e
vic
e
s.
A
m
o
ng
all
the
p
a
ra
meters th
at
affect the n
e
tw
ork
efficiency acc
u
racy, scalab
ilit
y, and pow
er consu
m
ption
ar
e ma
in ch
all
e
n
ges in the ro
uting of Mob
ile a
d
-
hoc netw
o
rks. T
he netw
o
rk
lifetime is dep
en
dent on the p
o
w
er effici
ency of the nodes i
n
the netw
o
rk. The
protoco
l
s hav
e
to provi
de th
e
ener
gy efficie
n
t route th
ro
ug
h inter
m
edi
ate
nod
es in t
he n
e
tw
ork. T
he trust
base
d
ro
utin
g
appr
oach
is
o
ne of th
e
best
mecha
n
is
ms t
o
esta
blis
h a
n
ener
gy effici
e
n
t route
betw
e
en
source
an
d d
e
s
tinatio
n. In thi
s
pap
er w
e
firs
t prop
ose th
e f
a
mily r
e
lati
ons
hip
bas
ed trust
mode
l a
nd t
h
en
prop
ose
a n
e
w
energy
effici
ent trust bas
e
d
routi
ng
pr
oto
c
ol to re
duc
e the ro
utin
g ov
e
r
hea
d, de
lay
a
nd
provi
des better
packet del
iver
y ratio that perform
s b
e
tter tha
n
the existin
g
routin
g protoc
ol
s.
Ke
y
w
ords
: Mo
bile
ad-h
o
c net
w
o
rks, Security, T
r
ust Manage
me
nt
Copy
right
©
2016 In
stitu
t
e o
f
Ad
van
ced
En
g
i
n
eerin
g and
Scien
ce. All
rig
h
t
s reser
ve
d
.
1. Introduc
tion
The MANET
s
are auton
omou
s sy
ste
m
of
portab
l
e wirele
ss
mobile n
ode
s that
comm
uni
cate
without any
spe
c
ific inf
r
a
s
tructu
re
o
r
ce
ntralized a
c
cess. Every mobile no
de in
the
netwo
rk
act
s
as a route
r
a
nd wo
rks a
s
an in
term
edi
ate node
bet
wee
n
so
urce
and de
stinati
on.
Many Rea
c
ti
ve, Proactive
and hybrid routing pr
otocols have be
e
n
prop
osed to make p
r
op
er
comm
uni
cati
on in the n
e
twork n
ode
s.
In MANETs,
node
com
m
u
n
icatio
n is d
e
pend
ent on t
h
e
mutual trust [
1
] [2] amon
g
the nod
es.
T
he
con
s
tr
ai
nts in th
e MA
NETs p
o
se ma
ny new resea
r
ch
chall
enge
s i
n
the routin
g, privacy, trust
an
d se
curity incl
udi
ng authe
ntication and
key
manag
eme
n
t among the n
o
des.
A Genetic al
gorithm b
a
se
d energy entropy
multipat
h routing
app
roachwas p
r
o
posedin
[16] to adjust
energy utilization
of indiv
i
dual no
de, calcul
ate
the
minimal en
ergy of node a
nd
drag
out th
e
lifetime and
energy chan
ge of the
sy
stem. Th
e e
nergy
savin
g
routin
g p
r
ot
ocol
s
have be
en d
e
sig
ned to i
m
prove th
e p
e
rform
a
n
c
e i
n
term
s of o
v
erhea
d in
ro
uting, end to
en
d
delay, PDR
of the netwo
rks and
con
s
umptio
n of the energy. Se
cu
rity is one of the main
chall
enge
s fo
r the practi
ca
l implementat
ion of ad-
ho
c netwo
rks, su
ch a
s
MANE
Ts o
r
Wi
rele
ss
sen
s
o
r
networks. Tra
d
itio
nally, functions that
drive
WSNs, such as mediu
m
accesscon
trol
(MAC) and ro
uting proto
c
ol
s, always a
ssume t
hat theoperating env
ironm
ent is trustworthy [15
]
.
This
assu
mpt
i
on is not
always ri
ght an
d
remote
environment
s a
r
e
alway
s
susce
p
tible to atta
cks
and a
r
e very
tough to p
r
otect. It is o
b
se
rved that
the ene
rgy i
nefficien
cy af
fects the
ove
r
all
netwo
rk
pe
rfo
r
man
c
e
and li
fetime. So we ca
n say tha
t
the insuffici
ent po
wer
of a nod
e lead
s to
link failures a
nd deg
rad
e
s t
he network p
e
rform
a
n
c
e.
The co
ncept of trust origi
n
ally taken fro
m
soci
al scie
nce
s
an
d is d
e
scrib
ed a
s
subje
c
tive
belief about the behavio
rs of a particul
a
r entit
y [3].
Tru
s
t manag
ement is introdu
ced [4] and
clarifie
d a
s
"trust ma
nage
ment provid
e
s
a uni
fie
d
approa
ch for spe
c
ifying a
nd interpretin
g
se
curity poli
c
ies, crede
ntials, and
rela
tionshi
ps
." T
he de
sign of
trust ba
sed
energy efficient
routing
proto
c
ol in this pa
p
e
r, first explai
ns
ab
out the
energy model
to
find the energy facto
r
a
nd
gives a overvi
ew of our trust based a
ppro
a
ch to be imp
l
emented.
The rem
a
inin
g part of the pape
r org
ani
zed a
s
follows: In sectio
n 2, the background an
d
related
work
is given. Section 3 descri
bes the
overview of trust and re
putatio
n model use
d
in
this meth
od.
Section
4
sho
w
s ou
r resea
r
ch fram
e
w
o
r
k and
propo
se
d ro
uting
prot
ocol. In S
e
cti
on
5,
extensive experim
ents and simulatio
n
s con
d
u
c
ted
in co
mpa
r
iso
n
with the
exi
s
ting p
r
oto
c
ol
s is
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 25
02-4
752
IJEECS
Vol.
2, No. 2, May 2016 : 417 –
425
418
pre
s
ente
d
, a
nd finally in
section
6, co
n
c
lu
sion
and
challen
ges en
cou
n
tere
d an
d future
scop
e is
pre
s
ente
d
.
2. Backg
rou
nd Work
For effici
ent
utilization of
battery po
we
r of
node
s in t
he net
work, variou
s p
o
wer
efficient
routing
me
ch
anism
s [5]-[1
1] have be
en
prop
osed. T
r
ust value
s
a
r
e used to
co
n
s
tru
c
t safe pa
ths
among the n
ode
s in the network. Co
nsidera
b
le am
o
unt of work is don
e on the power effici
en
t
routing p
r
oto
c
ol
s but not i
n
trust ba
se
d
approa
ch
. Authors [12] propo
sed a n
e
w ene
rgy efficient
and
secure SEER multipat
h routin
g protocol. Thi
s
protocol
updat
es each node
with remaining
energy on d
y
namic b
a
si
s for finding
the ap
pro
p
ri
ate path from
multiple choi
ce
s. The m
a
in
advantag
e of this kind of
approa
ch is
minimizi
ng th
e overhe
ad to maintain th
e route an
d
can
maximize the
efficiency an
d lifetime of other nod
es in
the netwo
rk.
Many autho
rs have discu
s
sed va
riou
s i
s
su
e
s
re
gardi
ng tru
s
t man
ageme
n
t in MANET'
s
and in
wi
rele
ss sensor net
works.
Th
e a
u
thors in [1
3]
have di
scu
ssed a
novel tru
s
t a
w
are routi
n
g
proto
c
ol that
use
s
dire
ct
trust and in
dire
ct trust. It has monito
ring compo
n
ent with sev
e
ral
metrics like data confidenti
a
lity,
data integrity, available ene
rgy,
network-a
ck, a
nd rep
u
tation
. A
TCLM [1
4], trust ba
se
d cross laye
r mo
del u
s
e
s
the
ACKs from DL layer
and
TCP to p
r
om
ote
trust an
d eli
m
inates th
e malicio
us n
o
des a
nd in
si
sts hig
h
ly trusted
route
from so
urce
to
destin
a
tion.
Tru
s
t mana
g
e
ment in MA
NETs i
s
n
eed
ed when n
e
w node
s join th
e network an
d wa
nts
to establi
s
h
a co
mmuni
ca
tion with
acceptable l
e
vel
of tru
s
t relat
i
onship
s
amo
ng them
selve
s
.
Trust management has applicabilit
y in many deci
si
on making
si
tuations incl
uding intrusi
o
n
detectio
n
, acce
ss
control, key manag
ement, authe
ntication an
d
for effective routing. Tru
s
t
manag
eme
n
t, includ
es tru
s
t establishme
n
t and tru
s
t revocatio
n
.
3. Trust Ev
aluation
In our e
a
rli
e
r wo
rk, we
prop
osed tru
s
t ba
sed
m
odel for
MANETs
usi
n
g
Family
relation
shi
p
b
a
se
d app
roa
c
h. The mi
sb
ehavior
of th
e node
s deg
rade
s the perf
o
rma
n
ce of the
netwo
rk,
so t
he tru
s
t mod
u
le used to
provide
se
cu
re commu
nication and
efficient routing
is
possibl
e. A mobile ad
-ho
c
netwo
rk fully
depe
nd
s on the co
-op
e
ration between n
ode
s for ro
uting
and fo
rwardin
g
. The
succe
ssful
delive
r
y of data
fro
m
sou
r
ce to
de
stination
will h
appe
n if all
th
e
node
s co-op
e
rate
well. T
he attacksare identifi
ed and solved by
intrusi
on detectio
n
,
se
cure
routing,
key manag
eme
n
t and tru
s
t manag
ement.T
his sectio
n d
i
scusse
s ab
o
u
t the differe
nt
ways of e
s
tab
lishin
g
trust b
e
twee
n nod
e
s
in mobile a
d
-ho
c
net
wo
rks.
Direc
t
Trus
t
The di
re
ct tru
s
t will
be
cal
c
ulate
d
by di
rect inte
ra
ctio
n bet
wee
n
im
mediate
neig
hbori
n
g
node
s i
n
the
netwo
rk a
s
shown in
Figu
re 1.
The
direct trust
can
log the
num
b
e
r of
su
cce
s
sful
packet tra
n
sf
ers, recomm
endatio
n and
misbe
havio
r detection. It is the mo
st widely u
s
ed t
r
ust
cal
c
ulatio
n method when t
here a
r
e n
o
p
r
e-esta
b
lishe
d infrast
r
u
c
tures an
d ce
ntra
lized
control.
Recomme
nd
ation Trus
t
There might
be some m
a
li
ciou
s n
ode
s,
whi
c
h
b
ehav
e differe
ntly with different
node
s. In
this
kind
of situation, the
dire
ct tru
s
t i
s
not
sufficien
t and
so th
e
re
comm
end
ation a
bout t
hat
particular node
will al
so be cons
idered
to cal
c
ulate trust. Here
the other
mutu
al nei
ghbors
will
sha
r
e
its tru
s
t table
with
the n
e
igh
b
o
r
node
s.
T
he trust
cal
c
ulation m
e
th
od i
s
kno
w
n
as
recomme
ndat
ion tru
s
t or i
n
dire
ct tru
s
t. There
are
som
e
pro
b
lem
s
i
n
this recom
m
endatio
n trust,
su
ch a
s
false
recomme
ndat
ion by the other nod
es d
ue
to maliciou
s
n
a
ture of network n
ode
s.
Trust Comp
uta
t
ion
In this sub se
ction
we di
scuss ab
out the diffe
rent tru
s
t com
putatio
ns u
s
e
d
in o
u
r work.
All trust val
u
e
s
comp
uted i
n
ou
r
schem
e ra
nge
s f
r
o
m
0 to
1. Ba
sed
on
the to
tal tru
s
t value
the
role/relationship will be
assigned
to the neighbor node. For
cal
c
ul
ating trust, we are using the
con
c
e
p
t of co
nvex hull whi
c
h give
s a va
lue that lies b
e
twee
n two fixed points.
Evaluation Warning : The document was created with Spire.PDF for Python.
IJEECS
ISSN:
2502-4
752
A Power Efficient Tru
s
t Based Se
cu
re
Routing Sch
e
m
e
for Mobile
Ad-Hoc …
(M
V Rathnam
ma)
419
Initial Trust
This t
r
u
s
t is calcul
ated
usi
ng the
pa
ram
e
ters,
battery
po
wer an
d
si
gnal
stre
ngth.
Thi
s
is
the ba
si
c
crit
eria fo
r
a n
o
d
e
to b
e
in
the
network. Thi
s
trust i
s
the
main fa
ctor to
de
cide
whet
her
to kee
p
the n
ode a
s
a
neig
hbor
or
not. Trust u
p
g
r
ad
ation al
so d
epe
nds
on thi
s
trust value. Init
ial
trust valu
e i
s
mainly u
s
ed
to re
du
ce t
he atta
cks b
y
the selfi
s
h
node
s
be
ca
use
of resou
r
ce
limitation.
∗
∗
In above
eq
u
a
tion
IT
re
pr
es
e
n
t
s
th
e in
itia
l tr
us
t,
BP
re
pre
s
ent
s the
battery po
we
r and
SS
rep
r
e
s
ent
s the signal
stre
ngt
h of neig
hbors of ne
w node
N
. The
and
r
e
pr
es
e
n
t
s th
e
variable
s
an
d
the summati
on sh
ould be
1. In our work
we have
taken 0.5, 0
.
5 for
and
r
e
spec
tively.
Behav
i
oral Trust
The beh
avioral trust is calculate
d
by di
rect inte
ra
ctio
n and expe
rience of one node to
anothe
r no
de
. The pa
ram
e
ters fo
r
cal
c
ul
ating be
havio
ral tru
s
t
will vary for
differe
nt node
s b
a
sed
on its level as mentioned in
table 1.
1
∗
In the above
equatio
n
BT
repre
s
e
n
ts the
behavio
ral trust,
l
rep
r
e
s
e
n
t
s the tru
s
t le
vel and
p
r
e
pr
es
en
ts
th
e
p
a
r
a
me
te
r o
f
N
o
de
N
. For exa
m
ple
if the node
le
vel is
l
= 2, th
e
p
1
and
p
2
of
node
N
will be taken as shown in Table
1.
1
∗
Recommendation Trust
The recomm
endation trust is calcula
t
ed from the mutual neighbors of a
n
y two
neighboring
nodes. All the mutual nei
ghbors will
s
hare their tru
s
t value or
opinion abou
t
a
particular nod
e to calculate
the recomme
ndation trust.
In the above
equation,
RT
repre
s
ents re
commendatio
n trust,
n
represents the nu
mber of
mutual neighbors and
t
i
represents the trust value shared by
i
th
mu
tual neighbor of node
N
.
Total Trus
t:
The total trus
t is calculated from Behav
ioral trust a
n
d recomme
ndation trust. The
total trust will be useful in
upgrading o
r
degrading
th
e trust level of a node. The total trust also
ranges from 0
to 1
.
∝
∗
∗
In the above equatio
n,
TT
repres
ent
s
total trus
t,
BT
repre
s
e
n
ts be
havioral trust
and
RT
rep
r
e
s
ent
s a recomme
ndat
ion tru
s
t of node
N
. Th
e variabl
es
an
d
sho
u
ld h
a
ve the values
su
ch that the summ
ation wi
ll be 1. In our work we co
nsider 0.7 an
d 0
.
3 for
and
res
p
ec
tively.
4. Algorithm and Propos
e
d
Work
This
se
ction
descri
b
e
s
a
b
out the
key id
ea of
o
u
r
pro
posed
wo
rk.
There a
r
e t
w
o differe
nt
pha
se
s name
l
y bootstrap
p
i
ng and up
gra
d
ing/do
wn
g
r
a
d
ing pha
se.
The boot
stra
pping ph
ase will
take pla
c
e
when a ne
w n
ode wants to
join the
network with
out
any previou
s
experi
e
n
c
e.
In
Bootstra
ppin
g
ph
ase initi
a
l tru
s
t is u
s
ed
for
t
r
u
s
t com
putation
.
The u
p
g
r
a
d
ing/do
wng
r
a
d
ing
pha
se will be
use
d
to updat
e the trust val
ue and relatio
n
of the neigh
bor no
de
s.
4.1. Boot Str
a
pping Phas
e
Whe
n
a ne
w node
want
s to join the netwo
rk, the
neighb
or
wil
l
che
ck
wh
ether a
n
y
mutual neig
h
bors are there or not. If any mutual
nei
ghbo
rs
are th
ere, the no
de
will req
u
e
s
t for
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 25
02-4
752
IJEECS
Vol.
2, No. 2, May 2016 : 417 –
425
420
recomme
ndat
ion trust fro
m
all other mut
ual neigh
bo
rs having relati
onship mo
re than or e
qual
to
“pa
r
ent”. Th
en the ne
w node will
be add
ed to
the netwo
rk one level
lesser tha
n
the
recomme
ndat
ion tru
s
t. If n
o
t, the initial trust
w
ill be calcul
ated an
d
the node
will be adde
d to the
netwo
rk with
least
privile
ge. The
wo
rking
of
boot
strap
p
ing
ph
ase i
s
d
e
scri
bed in foll
owing
algorith
m
:
Algorithm 1
(Bo
o
ts
tra
pping phase
):
// When ne
w node
want
s to be a neig
h
b
o
r
if (mutual nei
ghbo
r) {
cal
c
ulate recommen
dation
trust
RT
(N
)
add no
de
N
a
s
a neig
hbo
r (trust value
=
RT(
N
)/2
)
}
else {
cal
c
ulate initi
a
l trust
IT(N)
add
no
de
N
a
s
a neig
hbo
r (trust value
= 0)
}
4.2. Upgradi
ng/Do
w
n
g
r
a
d
ing Phase
Whe
n
a n
ode
want
s mo
re
privilege, it wi
ll se
n
d
an
up
date re
que
st
to its neig
hbo
r. First
initial tru
s
t is
cal
c
ulate
d
to
ensure
that t
he no
de i
s
h
a
ving sufficie
n
t re
sou
r
ces.
If the nod
e
has
sufficie
n
t re
source
s, total trust is
cal
c
ul
ated
from
be
havioral t
r
ust
and recomm
endatio
n tru
s
t. If
the total trust is greate
r
th
an thre
shol
d value 0.
75 th
en it is eligibl
e
for up-gra
d
a
tion. Otherwise,
it indicate
s th
e malici
o
u
s
b
ehavior; the
n
the node i
s
marked
as m
a
licio
us n
ode
by setting th
e
trust valu
e to
-1. If the
r
e
a
r
e n
o
sufficie
n
t re
sou
r
ce
s, but the
total
trust i
s
m
o
re
than 0.7
5
; th
en
the node i
s
n
o
t eligible for upgradatio
n. In this ca
se
the node
will
retain its ol
d
trust value. T
he
followin
g
algo
rithm explain
s
the actu
al
worki
ng of upg
rading/d
o
wn
grading p
h
a
s
e.
Algorithm 2
(Up
g
rading /
Do
w
n
gr
adin
g Phase):
// W
hen a trust
upgr
ade r
equ
e
s
t from node
N
calcul
ate in
itial
trust of node
N
[
IT(N)
]
if (
IT(N)>0.75
) {
calcul
ate total t
r
ust of node
N
[
TT(N)
]
if (
TT(N)>
0.75) {
upgr
ade nod
e
N
(trust value =
current trust *2)
}
else
mark nod
e
N
a
s
malicio
us no
de (trust value
=
-1)
}
else {
if (
TT(N)>
0.75) {
Don’t u
pgra
de
nod
e
N
(trust valu
e =
current trust)
}
else
mark nod
e
N
a
s
malicio
us no
de (trust value
=
-1)
}
4.3. Energ
y
Sa
v
i
ng Model
From tru
s
ted
energy
savi
ng persp
ectiv
e
, pow
e
r
a
w
are routing p
r
otocol PTSRP, based
on the
above
describe
d
trust meth
od f
o
r effici
ent
ut
ilization
of en
ergy a
nd u
s
e
s
hyb
r
id p
o
wer
saving
sche
me, whi
c
h is
more b
a
lan
c
ed and
se
cu
re is propo
se
d. It provides better sh
ari
n
g of
netwo
rk
reso
urces a
nd m
a
intain
s efficient powe
r
sa
ving. This is achi
eved by isolatin
g the bad
node
s an
d de
legating p
a
rt of the trust ca
lculat
io
ns to the se
nde
rs a
s
ba
se statio
ns.
Evaluation Warning : The document was created with Spire.PDF for Python.
IJEECS
ISSN:
2502-4
752
A Power Efficient Tru
s
t Based Se
cu
re
Routing Sch
e
m
e
for Mobile
Ad-Hoc …
(M
V Rathnam
ma)
421
To illustrat
e
the benefits
of PTSRP,
we will sh
o
w
the calcul
ations for a
bov
e approa
ch
with
respe
c
t to th
e po
wer. T
h
e
overhe
ad i
s
cla
ssifie
d
to two different
parts, th
e re
p
o
rts
se
nt by the
node
s an
d ce
ntral re
port sent by the BS.
Energy 1 =
1
(1)
Her
e
M - Size of the messag
e, per neig
hbo
r in
bytes.
N - Total no
d
e
s in the net
work.
h- Averag
e n
u
mbe
r
of hop
s from no
de to BS.
AN- Active ne
ighbo
rs of a n
ode.
Rx energy- Energy to re
cei
v
e one byte
Tx energy- E
nergy to tran
smit one byte
As pe
r th
e
BS cent
ral
report, it
con
s
ist
s
of
me
ssag
es from
all mali
ciou
s nod
es an
d
is
broa
dcaste
d to all the node
s for every time peri
od t is calculated a
s
follows
Energy 2 =
(2)
Whe
r
e Ml is t
he numb
e
r of
maliciou
s
no
des.
Assum
e
tx=E and normalizin
g Rx we obtai
n,
1
1
1
(3)
2
1
(4)
If the average
time interval of droppi
ng p
a
ckets i
s
td
2
1
(5)
w
h
er
e
is the pa
cket size in bytes
Packet follows the full dupl
ex commu
nication
and divi
ding by Rx to norm
a
lize, we obtain
2
1
(6)
The ene
rgy saving ca
n be
obtaine
d for the frequ
ent p
e
riod
s
is calculated a
s
/
1
2
⁄
(7)
Es is con
s
ide
r
ed a
s
the en
ergy savin
g
s
for the
PTSRP and all the values a
r
e filtered
re
sults o
f
trust from the
previou
s
se
ction algorith
m
s.
5. Results a
nd Analy
s
is
I
n
t
h
is
se
ct
io
n,
we
de
sign
som
e
si
mula
t
i
on test exp
e
rime
nts of P
T
SRP protocol usi
n
g
Network
Simulator-2. In this
s
i
mulation t
e
s
t
ex
p
e
rim
e
nts, we
simul
a
te an
d
com
p
are
ene
rgy
cost
and total re
ceiving data p
a
ckets a
gain
s
t the ot
her
prop
osed pro
t
ocol
s like S
A
ODV [17] and
TRRP [18].
5.1 Performa
nce Ev
aluation
Total Throug
hput:
The total numbe
r of packet
s
re
cei
v
ed per unit time.
Total O
v
e
r
he
ad:
Total n
u
m
ber
of ro
uting contro
l pa
ckets tran
smi
tted at time
t
by all the n
o
d
e
s in
the netwo
rk.
P
a
ck
et
D
e
liv
e
r
y
Rat
i
o:
It i
s
the ratio
of to
tal numb
e
r of
pa
ckets
su
ccessfully deliv
ered
to the
to
ta
l
numbe
r of pa
ckets sent.
P
a
ck
et
lat
e
n
c
y:
the total time elap
se
d since
a data
p
a
cket is t
r
an
smitted to time the data p
a
cket
rea
c
he
d the d
e
stinatio
n.
The simulatio
n
s are
condu
cted
to exami
ne
the pe
rformance by
ad
ding
se
cu
rity. He
re PTS
R
P
is
comp
ared to SAODV and
TRRP.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 25
02-4
752
IJEECS
Vol.
2, No. 2, May 2016 : 417 –
425
422
Table 1. Simulation Para
meters
Number of
Node
s
100
Topolog
y dimension
1000m x
1000m
Radio range
250m
Node pause time
s
0-40s
Traffic Pattern
FTP/T
CP`
Maximum node s
peed
1-20m/s
Source-destination pairs
20
In our scen
ario, simulation
s co
ndu
cted t
o
examine th
e perfo
rman
ce by adding
se
curity
to the routing
protocols. We comp
are o
u
r propo
sed
model with e
x
isting two ro
uting proto
c
ol
s
and obtai
n b
e
tter re
sults.
Simulation p
a
ram
e
ters
are given in th
e table, and
a malici
o
u
s
n
ode
rand
omly dro
p
s data p
a
ckets and
can
be dete
c
ted
durin
g format
ion of netwo
rk topolo
g
y. Here
the drop
ping i
s
in the scale
of 20% to 50% and eac
h simulation tim
e
is 600
s to collect the out
put
data.
0
100
200
3
0
0
4
00
5
0
0
6
0
0
700
800
0
5
10
15
20
T
o
t
a
l
T
h
ro
ughpu
t
(
P
a
cket
s
x10
3
)
Ti
m
e
(
s
)
SA
OD
V
T
RRP
PT
S
R
P
Figure 1. Total Thro
ugh
pu
t in the prese
n
ce of 5 mali
ciou
s no
de
s
Figure 1 re
prese
n
ts the through
put of the three p
r
oto
c
ol
s und
er five malici
ous
n
ode
s out
of 50. All the
routing
proto
c
ol
s a
r
e d
e
livering th
e pa
ckets to th
e de
stination
s
d
u
e to less n
u
m
ber
of mali
ciou
s
node
s.
Ho
we
ver, ou
r p
r
o
p
o
se
d met
hod
outpe
rforms
the othe
rs, h
ence the
efficient
res
u
lts.
0
100
200
300
400
5
0
0
6
00
7
0
0
800
0
2
4
6
8
10
12
14
16
18
20
T
o
t
a
l T
h
roughput
(Packet
sx10
3
)
Ti
m
e
(
s
)
SA
OD
V
TR
RP
PTSR
P
Figure 2. Total Thro
ugh
pu
t in the prese
n
ce of 10 mal
i
ciou
s no
de
s
Evaluation Warning : The document was created with Spire.PDF for Python.
IJEECS
ISSN:
2502-4
752
A Power Efficient Tru
s
t Based Se
cu
re
Routing Sch
e
m
e
for Mobile
Ad-Hoc …
(M
V Rathnam
ma)
423
If the num
ber of mali
ciou
s nod
es in
cre
a
se
s f
r
om
5
to 10, a
nd to
20, a
s
sh
own in t
h
e
Figure 2 and
Figure
3, we can
ob
serv
e the pa
cket delivery of T
RRP a
nd SAODV de
crea
ses
proportionally, whereas PT
SRP st
ill deli
v
ers the packets efficientl
y. SAODV stops delivering of
packet
s
at ti
me t=5
40 in
the 30% to
4
0
% malici
o
u
s
node
s.
Due
to the he
avy packet d
r
o
p
, the
con
n
e
c
tion
wi
ll be
timed
ou
t and
ne
w
ro
ute di
scovery
will
be
initiat
ed a
gain.
Even m
o
re
n
u
m
ber
of malici
o
u
s
node
s i
n
the net
wo
rk,
PTSRP di
scovers th
e t
r
ust
w
o
r
thy ro
utes a
nd
re
sults
su
ccessful pa
cket delivery.
0
100
200
300
400
5
0
0
6
00
7
0
0
800
0
2
4
6
8
10
12
14
16
18
20
T
o
t
a
l T
h
roughput
(Packet
sx10
3
)
Ti
m
e
(
s
)
B
C
D
Figure 3. Total Thro
ugh
pu
t in the prese
n
ce of 20 mal
i
ciou
s no
de
s.
Figure 4
rep
r
ese
n
ts the to
tal overhe
ad
of
SAODV, T
RRP
and
pro
posed PTS
R
P. From
the analysi
s
of the results, PTSRP has t
he le
ss ove
r
he
ad
than the remainin
g ro
uting
approa
che
s
.
The
ba
sic re
aso
n
b
ehin
d
t
h
is i
s
th
at the
PTSRP
dete
c
t the
mali
cio
u
s
nod
es u
s
i
ng
trust ba
sed
mech
ani
sm a
nd avoid
s
those n
ode
s fr
om the routin
g. SAODV tend
s to wait and
time out often.
6
9
12
15
1
8
21
24
2
7
30
9
10
11
12
13
14
15
16
17
18
19
20
T
o
t
a
l O
v
e
r
he
ad
(
P
a
c
k
e
t
s
x1
00
0)
Num
b
er
o
f
M
a
licio
us
Nod
e
s
SAOD
V
T
RRP
PT
SR
P
Figure 4. Total overhe
ad i
n
the pre
s
en
ce of maliciou
s
node
s
Figure 5
sho
w
s th
e PDR
of PTSRP at differ
ent
spe
eds
com
p
a
r
e
d
to the T
R
RP and
SAODV. PTSRP ch
oo
se
s the more
reli
able route
s
b
y
avoiding th
e more m
a
licious n
ode
s a
nd
increa
se
s the
efficien
cy. The spee
d in
creases f
r
om 1
.
3 to 2.6 m/s,
even thou
gh
link b
r
e
a
kag
e
s
may red
u
ce t
he pa
cket de
livery ratio, the nod
es
are
more li
kely to find the av
ailable p
a
irs
to
forwa
r
d the p
a
ckets.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 25
02-4
752
IJEECS
Vol.
2, No. 2, May 2016 : 417 –
425
424
02
4
6
8
1
0
0.4
0
0.4
5
0.5
0
0.5
5
0.6
0
0.6
5
0.7
0
0.7
5
0.8
0
0.8
5
0.9
0
Packe
t
D
e
l
i
v
ery
R
a
t
i
o (
%
)
M
a
x
Sp
eed
(
m
/s)
SAODV
TRR
P
PT
S
R
P
Figure 5. Packet Delive
r
y Ratio at different spe
e
d
s
As the
maxi
mum
spe
ed i
n
crea
se
s fro
m
2.5 to
10
m/s,
the
li
nk brea
ka
ge
i
s
main cau
s
e
that redu
ce
s the pa
cket del
ivery ratio. P
DR d
e
crea
se
s as the m
a
ximum sp
eed i
n
crea
se
s.
6. Conclusio
n
s and Fu
tur
e
Work
From th
e re
sults of si
mula
tions, we su
mmari
ze th
e
contri
bution
o
f
this re
se
arch; PTSRP
is
suitable
fo
r the
se
cu
re
routin
g
with
truste
d valu
es i
n
MA
NETs d
ue to
it'
s
con
s
id
era
b
l
e
accuracy, av
erag
e path l
ength an
d m
oderate ene
rgy con
s
um
ption. This p
a
per p
r
op
ose
d
a
method f
o
r t
r
ust
calculation, an
d the
tru
s
t me
ch
anism
integ
r
ated
with th
e efficie
n
t p
o
we
r
utilization model and gives the better
results
than widely used AODV routing
scheme.
T
h
e
proposed PT
SRP outperf
orm
s
the
existing
routing
protocol
s in the perf
orm
ance. Still it is
possibl
e to improve the en
e
r
gy saving
scheme
by re
du
cing
cal
c
ulati
on overh
ead
of trust.
Referen
ces
[1]
JH Cho, A S
w
a
m
i, and IR
Che
n
. “A Surve
y
o
n
T
r
ust Manag
ement for Mo
bi
le Ad
Hoc N
e
tw
o
r
ks”.
IE
EE
Co
mmun
icati
o
ns Surveys & T
u
toria
l
s
. 201
1; 13(4).
[2]
L Capr
a. “
T
o
w
a
rd a Hu
ma
n T
r
ust Model
for Mobile A
d
-hoc N
e
tw
orks
”. Proc. 2nd
UK-UbiN
e
t
W
o
rkshop, Ca
mbridg
e Un
iver
sit
y
, Cambr
i
d
g
e
, UK. 2004.
[3]
KS Cook (editor).
T
r
ust in Society
. Russell Sa
ge F
oun
dati
on
Series o
n
T
r
ust, Ne
w
Y
o
rk. 20
03; 2.
[4]
M Blaz
e, J F
e
i
gen
ba
um a
nd
J Lac
y. “
D
e
ce
n
t
rali
z
e
d
T
r
ust
Mana
ge
me
nt
”. Proc. IEEE S
y
mposium o
n
Securit
y
an
d Privac
y
.
1
9
9
6
: 164 - 173.
[5]
D F
eng a
nd Y
Z
hu. “An Improved AODV R
outin
g Protoco
l
Based o
n
Re
maini
ng P
o
w
e
r
and F
a
me”
.
Electron
ic Co
mputer T
e
chn
o
l
o
gy
. 2009: 1
17-
121.
[6]
V Rod
o
p
l
u a
n
d
T
H
Meng. “
M
inimum
Ener
g
y
Mob
ile
W
i
reless
Net
w
ork
s
”.
IEEE Journal on Selected
Areas in C
o
mmu
n
ic
ations.
1
999; 17(
8): 133
3-13
43.
[7]
PJ W
an, G Cali
nesc
u
, XY
Li a
nd O F
r
ied
e
r. “Minim
u
m-ener
g
y
bro
adcasti
ng i
n
s
t
atic ad
ho
c
net
w
o
rks”.
W
i
rel. Netw
. Journal
. 200
2; 8(6): 607-
617.
[8]
I Park, T
H
Meng
a
nd I
Pu
. “Blocki
ng
E
x
pan
din
g
Rin
g
Searc
h
A
l
gor
ithm for Ef
fi
ci
ent En
erg
y
Cons
umptio
n i
n
Mobi
le Ad H
o
c Net
w
orks”.
IFIP W
O
NS
. 2006: 191-
19
5.
[9]
LM Fee
n
e
y
an
d M
Nilss
on. “I
nvestig
a
tin
g
th
e En
erg
y
Co
ns
umptio
n of
a W
i
reless
N
e
t
w
or
k Interface
i
n
Ad hoc Net
w
o
r
k Environm
ent”
.
IEEE IN
FOC
O
M
. 2001; 3: 1548-
155
7.
[10]
N Sumathi a
n
d
AS
T
hanam
ani. “Eval
uatio
n of Energ
y
E
f
fi
cient Reactiv
e
Routi
ng roto
cols in Qo
S
Enab
led R
outi
ng for MANET
S
”.
Int. Journal
of Computer A
pplic
atio
ns
. 20
11; 14: 10-
14.
[11]
A S
y
arif a
nd R
F
Sari. “Performance An
al
ysi
s
of
AODV-UI
Routi
ng Protoc
ol
w
i
th En
erg
y
Cons
umptio
n
Improveme
n
t u
nder M
obi
lit
y
Mode
ls in
H
y
b
r
id Ad
hoc N
e
tw
o
r
k”.
Int. Jou
r
nal
on C
o
mp
uter Scie
nce
and En
gi
neer
in
g (IJCSE).
201
1; 3: 2904-
291
8.
[12]
N Nasser an
d
Y Chen. “SEEM: secure an
d ener
g
y
-e
ffici
ent multip
ath routin
g protoco
l
for
w
i
rel
e
ss
sensor n
e
t
w
ork
s
”.
Comp
uter C
o
mmunic
a
tio
n
s
. 2007; 30(
11-1
2
): 2401
–2
412.
[13]
T
heodore Z
a
h
a
ria
d
is, Pa
na
gi
otis T
r
akadas
,
Hel
e
n
Le
lig
ou
, Pana
giotis
K
a
rkazis. “
I
m
p
l
ementi
ng a
T
r
ust-Aw
are Routin
g Protoco
l
in W
i
reless Se
nsor No
des
”. D
e
SE 201
0, Lon
don UK. 20
10.
[14]
Rah
hal
HA, A
li IA, Sha
h
e
e
n
SI. "A nove
l
T
r
ust-Based Cross-La
ye
r
Mode
l for W
i
r
e
less S
ens
or
Net
w
orks". 28th
Natio
nal R
a
d
i
o Scie
nce Co
n
f
erence (N
RSC
)
. 2011; 1(1
0
).
Evaluation Warning : The document was created with Spire.PDF for Python.
IJEECS
ISSN:
2502-4
752
A Power Efficient Tru
s
t Based Se
cu
re
Routing Sch
e
m
e
for Mobile
Ad-Hoc …
(M
V Rathnam
ma)
425
[15]
G
T
heodorak
o
pou
los an
d J
Baras. “On trus
t models a
nd trust evalu
a
tion metrics
for ad ho
c
net
w
o
rks”.
IEEE J. Sel. Areas Commun
. 20
0
6
; 24(2): 31
8 – 328.
[16]
B Sun, C
Gui,
and
P L
i
u. “
En
ergy Entr
opy M
u
ltip
ath R
outin
g Opti
mi
z
a
ti
on
Algorit
h
m
i
n
M
A
NET
bas
e
d
on GA
”. Proc.
Fifth IEEE
Int.
Conf. on Bio-I
n
spir
ed Co
m
p
uting: T
heories
and App
licati
o
ns (BIC-T
A).
201
0: 943-
947.
[17]
MG Z
apata an
d N Asok
an. “S
ecuri
ng Ad h
o
c
Routin
g Proto
c
ols”. W
i
Se. 20
02.
[18]
Neel
aka
n
d
an
S, Anan
d JG. "
T
ru
st based
op
timal ro
utin
g i
n
MANET
'
s".
Emer
gin
g
tre
nds
in
Electric
a
l
and C
o
mput
er T
e
chno
logy (IC
E
T
E
CT
. 2011: 115
0, 115
6.
Evaluation Warning : The document was created with Spire.PDF for Python.