IMPROVING DATA TRANSMISSION IN THE VANET USING MULTI-CRITERIA DECISION MAKING METHOD BASED ON FUZZY LOGIC

In e na ional Jou nal in Founda ions o Compu e Science & Technology (IJFCST), Vol. 3, No.6, No embe 2013
DOI:10.5121/ij cs .2013.3606 61
IMPROVING DATA TRANSMISSION IN THE VANET
USING MULTI-CRITERIA DECISION MAKING
METHOD BASED ON FUZZY LOGIC
Ja ad Badali 1 Mo aza Mokh a i Naza lou2 pa in a u 3
1,2,3 Depa men o Compu e , maku B anch, Islamic Azad Uni e si y, maku, I an
ABSTRACT
In ehicula ad-hoc ne wo ks he packe s a e sen using mul i-hop me hods and he ecei ing limi o a
message is g adually ex ended, bu he exponen ial inc emen o he numbe o nodes e-b oadcas ing a
message esul s in b oadcas s o m p oblem in da a b oadcas ing in his case. Some cha ac e is ics like
high speed o nodes, apid opological changes and epe i i e discon inui ies ha e made i di icul o
design an e icien b oadcas ing p o ocol o hese ne wo ks.
We ha e o e ed a no el uzzy me hod based on mul i-c i e ia decision-making (MCDM) o p io i izing he
ehicles in selec ion o he mos p ope neighbo o b oadcas da a in his pape . Wi h using his uzzy
me hod, he mos p ope ehicles pa icipa e in da a b oadcas ing. The esul s o simula ion using NS show
ha because o selec ing he neighbo ing ehicles wi h high p io i y in da a b oadcas ing, he speed o
sending he packs is inc eased and he ne wo k load is conside ably dec eased. This me hod also
conside ably dec eases b oadcas ing a ic.
KEYWORDS
VANET, Da a Dissemina ion, B oadcas S o m, Fuzzy Decision Making
1. INTRODUCTION
VANETs a e subse o MANET known as new gene a ion o ad-hoc ne wo ks [4,6]. In o de o
es ablish he communica ion VANET, each ehicle is as a node which can ac bo h as ecei e
and sende and he eby b oadcas di e en in o ma ion be ween he ehicles. In hese ne wo ks,
he ehicles a e equipped wi h wi eless e minals wi h s anda ds like DSRC wi h sending limi
ex endable up o 1000m. Because o limi ed adio ange o each node in VANETs, i is equi ed
o e-b oadcas he ecei ed b oadcas ed message o he neighbo s. This ype o sending is called
mul i-hop and equi es ou ing algo i hms. Rou ing in VANETs is e y complica ed and di icul
because o some cha ac e is ics like high dynamism, high speed o ehicles and high b oadcas ing
scale o in o ma ion and he old ou ing me hods a e no su icien in hese ne wo ks.
In mul i-hop sending, he ecei ed limi o a message is g adually ex ended; bu in his case he
exponen ial inc easing o he numbe o nodes e-b oadcas ing he message b ings he p oblem o
b oadcas s o m in b oadcas ing o in o ma ion. The ollowing cases can be men ioned among he
impo an cha ac e is ics o VANETs [2]:
 VANETs ha e dynamic opology bu hey a e geog aphically limi ed.
 These ne wo ks ha e po en ially wide scale in b oadcas ing o in o ma ion
 They mos ly su e om desul o iness
In e na ional Jou nal in Founda ions o Compu e Science & Technology (IJFCST), Vol. 3, No.6, No embe 2013
62
 The densi y o ne wo k is a iable and is a unc ion o a ic o ehicles
 Topology o he ne wo k highly depends on he ea men o he d i e
 VANETs ha e e y low diame e in compa ison wi h public ne wo ks
 Exis ence o obs acles in u ban en i onmen s in he ne wo ks can dis u b ou ing
The ehicle- o- ehicle communica ions in VANET ne wo ks, da a and message exchange
be ween he nodes ( ehicle) is done in wo ways as ollows:
 Vehicle- o- ehicle communica ions (V2V)
 Vehicle- o- oad side equipmen communica ions (V2R)
Many s udies a e pe o med on V2V and V2R communica ions and i has been de e mined ha
VANETs wi h V2V-based da a communica ion ha e some p io i ies o e V2R-based VANETs
[14] and ha is why we ha e concen a ed on V2V communica ions. Conside ing he impo ance
o c ea ing a dedica ed equency band o wi eless communica ion in VANET ne wo ks, he
Fede al Communica ions Commission o he USA (FCC) has dedica ed a 75MHz bandwid h in
5.9GHz o Dedica ed Sho -Range Communica ions. The channel dedica ion me hod in his band
is b ie ly shown in igu e 1.The equency limi is di ided in o 7 channels wi h 10MHz
bandwid h; ou o which 1 channel is mainly dedica ed o inc ease he sa e y coe icien o oads
and he emained 6 channels a e dedica ed o se ice-wel a e applica ions [3].
Figu e1. Me hod o Dedica ing F equency Band & Di ision o I s Channels o DSRC/WAVE [14]
The messages exchanged in V2V ne wo ks o sa e y applica ion a e di ided in o wo ca ego ies
in e ms o hei na u e and schedule o b oadcas ing [15]:
Pe iodic messages al e na i ely b oadcas ed by each ehicle: These pe iodic messages con ain he
mo ional s a us o each ehicle in he ou e. By analyzing hese messages in each ehicle, we can
ob ain comple e da a abou he s a us o he ou . In o he wo ds, hese messages a e called
beacon.
Eme gency messages (e en d i en): hese messages a e issued om ime o ime and in case o
haza dous e en s in he ou in o de o wa n o he ehicles. In case o ecei ing eme gency
messages, we can p e en om mo e damages by making p ope decisions wi hou losing he
ime.
In e na ional Jou nal in Founda ions o Compu e Science & Technology (IJFCST), Vol. 3, No.6, No embe 2013
63
On he whole, some p oblems like weakening o he ecei ed signal le el, sa u a ion o channel
loading, high mobili y ne wo k nodes and b oadcas s o m should be o e came in o de o ha e a
eliable and sa e communica ion in VANETs. The p oblem o B oadcas S o m in hese ne wo ks
occu s because o some cha ac e is ics like high mobili y o nodes, high scale and s uc u e in
b oadcas ing and we y o sol e hese p oblems in his pape .
In his pape , a uzzy me hod is p oposed o selec he ehicles pa icipa ing in b oadcas ing,
which is compa ible wi h special cha ac e is ics o VANE ne wo ks. As uzzy in e ence sys ems
and uzzy heo ies in gene al ha e an impo an ole in op imiza ion and p io i iza ion p ocess,
his pape wan s o desc ibe he applica ion o uzzy decision-making in imp o emen o
b oadcas s o m. The uzzy heo y in gene al and uzzy se s will be desc ibed in sec ion3 o ge
mo e amilia wi h uzzy heo y.
2. RELATED WORK
Di e en me hods a e desc ibed in o de o sol e he p oblem o B oadcas S o m in VANETs.
One o he me hods used o sol e his p oblem in MANET Ne wo ks is ha i limi s he numbe
o s eps cou sed in one mul i-hop sending by de ining some h esholds [9]. Some changes a e
made in hese ne wo ks o gene alize hese me hods o VANETs. Fo example, om he ecei ed
signal s eng h (RSS) o a message, one pa is used as a c i e ion o decision-making o e-
b oadcas ing as ollows:
min
max
1ij
ij
MIN
RSS RSS
pRSS RSS

   (1)
In Equa ion (1), RSSmin is he minimum ecei able signal le el and RSSmax is he maximum sen
signal s eng h. Fo each message eaching om node i o j, he ecei ed signal s eng h is
calcula ed and placed in RSSij. The pij a iable is e-b oadcas ing p obabili y o he ecei ed
message. The highe he le el o ecei ed signal, he close i is o he sou ce node and he
ecei ed message will be e-b oadcas ed wi h less p obabili y in his case. This will inally esul
in educ ion o b oadcas s o m nea he sende [12].
In ano he me hod, all message- ecei ing nodes don’ e-b oadcas i and only he ecei e in
ex eme poin o he des ina ion should e-b oadcas i o o he nodes. O cou se, we will need a
communica ion wi h loca ion- inding equipmen such as GPS in o de o implemen his p o ocol.
Reques o B oadcas /Clea o B oadcas (RTB/CTB) hand-shale messages a e used in his
p o ocol [7].
In ano he me hod, epea ing me hod is used o send eme gency messages. In his de ini ion i an
acciden occu s, eme gency wa ning messages (EWM) a e sen wi h a high a e and hen he a e
o sending is dec eased [13].As a simple me hod o disco e b oadcas s o m, each node
in es iga es he sequence numbe o i s ecei ed packs and calcula es he ecei ing a e o packs
based on i . I he ecei ing a e is dec eased in he p esen ime in compa ison wi h he p e ious
imes, hen he p oblem o hidden node and collision o ames occu in ne wo k. So, i is be e o
inc ease he wai ing ime o occupy he channel. This is done by inc easing he con en ion
windows. In his case, by dec easing he compe i ion o sending and dec easing o ne wo k
loading, he a e o collision is dec eased and ecei ing a e o b oadcas ing ames is inc eased.
He e he es ima ed eceip a e in communica ion be ween wo nodes in he ecei e is calcula ed
acco ding o he ollowing accumula ed equa ion:
In e na ional Jou nal in Founda ions o Compu e Science & Technology (IJFCST), Vol. 3, No.6, No embe 2013
64
1
Re Re (1 ) Re
Es cp Es cp Sample cp
 

     (2)
Alpha coe icien in equa ion 2 is used o change he impo an o ecei ing a e in he pas in
compa ison wi h he p esen in o de o calcula e he es ima ed eceip a e. The a e age local
eceip a e is calcula ed om all sending nodes o he ecei e acco ding o equa ion 3 [1].
Re
Re
Es cp
Local cpRa e
Nodes
 
(3)
Ano he solu ion o dec easing he looding a ic o disco e ing he ou e is op imal ou ing
me hod. Clus e ing ou ing is one o he ou ing me hods which is called CBRP (Clus e Based
Rou ing P o ocol) and dec eases b oadcas s o m. In clus e ing me hods, he mo ing nodes a e
di ided in o di e en ca ego ies and a e placed nex o each o he in one clus e acco ding o
special ules. In one clus e , he nodes can ha e di e en oles such as head clus e , ga e and
o dina y nodes. In his me hod, he clus e s a e de e mined and hen some c i e ia like node's
a ibu e, clus e and head clus e a e p o ided. Ei c i e ion is conside ed o de e mine head clus e
which is calcula ed as ollows:
 
0,
max
d
E e T
i

  (4)
is he ime o p esence in oad and d is speed de ia ion a e. Each node wi h bigge E is selec ed
as head clus e . Tha is, a ehicle mo ing o a longe ime in he oad wi h lowe speed de ia ion
is selec ed. All nodes b oadcas Hello messages al e na i ely wi h speci ic in e als. The
messages con ain all da a o he neighbo ing able and he adjacen clus e able.
3. FUZZY THEORY
Fuzzy sys ems a e based on ules o knowledge. The hea o a uzzy sys em is a da abase which
is o med o uzzy i - hen, conclusion and decision-making ules. Fuzzy decision-making is an
ac ion o decision-making, some wo ds o which a e de e mined using con inuous unc ions. He e
some ad an ages o uzzy logic a e men ioned [11].
- Simple concep ual unde s anding
- Flexibili y
- P oduc ion o complex non-linea ma hema ical me hods
- Logical and a ional jus i iabili y
- Possibili y o designing based on expe imen al heo ies o expe s
Unlike i s appa en simila i y wi h p obabili ies, Fuzzy heo y is a heo y independen o
p obabili ies. Using o uzzy sys ems in wo ollowing s a es is e y impo an :
1- In s udying o e y complex sys ems
2- Unde condi ions whe e access o an app oxima e bu e y apid me hod is p e e ed
3.1. Fuzzy Se s
Global se in uzzy se s is all accessible da a abou a p oblem. Membe ship unc ions a e
exp esses wi h 0 o 1 in classic se s and his means ha : is he conside ed componen a membe
In e na ional Jou nal in Founda ions o Compu e Science & Technology (IJFCST), Vol. 3, No.6, No embe 2013
65
o he se o no . Bu membe ship unc ions in uzzy se s alloca e an amoun be ween 1 and 0 o
hemsel es.
I he membe s o global se a e discon inuous in uzzy heo y, he uzzy heo y is de ined as
ollows:
(5)
Fuzzi ica ion S ages o se s a e done in ou s ages:
1- Fi s , he minimum and maximum amoun s o global se a e de e mined.
2- The basic e bal amoun s a e de e mined. Fo example, we can use e bal amoun s such as:
Ve y Low, Low, Medium, Lowe Medium, Highe Medium, High and Ve y High.
3- The space be ween minimum and maximum o global se is di ided in o p ope sec ions.
This di ision can be linea o non-linea . The numbe o hese sec ions can be o canno
be equal o he numbe o e bal amoun s.
4- P ope membe ship unc ions o membe s o global se a e de ine o ep esen he e bal
amoun .
The bes me hod o ob aining he membe ship unc ions is using o ma hema ical equa ions,
because he neu ali y is kep by using his me hod. Se e al ma hema ical unc ions a e sugges ed
o his pu pose, among which h ee unc ions including T iangula , Bell-Shape and Gaussian a e
used mo e han o he s.
The main cha ac e is ics o Gaussian and Bell-Shape unc ions is ha hey a e close o he
hinking me hod o human and he ad an ages o iangula unc ion is ha we can ci e mo e
heo e ic easoning o p o e he heo ies[8]. The e o e, we will use his me hod in he p esen
pape .
The ollowing equa ion should be also used o de e mine he membe ship deg ee o membe s in
global se in a uzzy quan i y which ollows a iangula unc ion:
0 ,
,
( , , )
,
0 ,
x a
x a
a x c
c a
A c a b b x
c x b
b c x b


 

  
 












(6)
3.2. Fuzzy Decision-Making
Mul i-C i e ia Decision Making (MCDM) is he mos common modeling me hod in decision-
making p oblems which ies o model he decision-making p oblem unde condi ion whe e he
numbe o objec ions and a ibu es o decision-making is mo e han one. The aim o MCDM
echnique is o design and help decision-making p ocess o dis inguish he mos p ope solu ion
acco ding o he will o he decision make abou he p oblem. MCDM is di ided in o wo
ca ego ies: MODM (Mul i-Objec i e Decision Making) and MADM (Mul i-A ibu able Decision
Making) [5]. Me hods o sol ing mul i-objec i e decisions-making p oblem can be also easily
used o uzzy numbe s such ha calcula ions a e done on membe ship unc ions o uzzy se s

In e na ional Jou nal in Founda ions o Compu e Science & Technology (IJFCST), Vol. 3, No.6, No embe 2013
66
ins ead o nume ical amoun s. In his s a e o uzzy se s, he c i e ia showing he impo ance o
objec i es (weigh o c i e ia) and uzzy se s o op imali y show also he a e o op imali y.
3.3. Mul i-C i e ia Decision-Making wi h Mul i-Objec i e Decision-Making View
(MODM)
One o he me hods o sol ing mul i-c i e ia decision-making p oblems is analyzing i as a mul i-
objec i e p oblem, such ha each c i e ion is conside ed as an objec i e whe e we wan o each
[10]. In mul i-objec i e decision-making, we wan o a se o impo an objec i es by selec ing he
bes al e na i e. In hese issues, we conside ha we ha e a se o objec i es as O= {o1, o2, o3, ….,
on} and a se o al e na i es A= {A1, A2, A3, … , An}. We decide o each hese objec i es. In
o he wo ds, we wan o each o1, o2,… and on. I we show he decision wi h D, so we ha e:
D = o1^o2^o3^…^on
Tha is, o1, o2,… and on objec i es a e eached. The a e o eaching o j h objec i e is shown as
Dij. In o he wo d, he a e o eaching o j h objec i e is equal o Dij. I he impo ance o j h
objec i e is Bj, hen selec ing i h al e na i e will ul ill his need as aij.
And using maximum me hod o ga he ing we can each he Dij.
max( , )
D B a
ij ij
 (7)
Conside ing he combina ion (and) in eaching he se o objec i es, he a e o decision o
i hal e na i e is ob ained as ollows:
   
( ) min min max( , )
1 1
n n
D A D B a
i ij ij
j j
 
  (8)
4. RESEARCH METHOD
Ou P oposed me hod in his pape is p io i iza ion me hod o da a b oadcas ing in u ban VANE
which ac s in acco dance wi h Mul i-C i e ia Decision-Making (MCDM) wi h Mul i-Objec i e
Decision-Making me hod (MODM) in uzzy heo y.
In he sugges ed me hod, we assume ha he da a abou speed, he dis ance be ween he nodes
and he numbe o neighbo s o each node a e kep in a able. We ha e used GPS o de e mine he
loca ion and dis ance be ween he ehicles and we ha e also aken ad an age o beacon messages
o calcula e he numbe o neighbo s. The cha ac e is ics o he sugges ed p o ocol a e as ollows:
 Fuzzy Va iable o Vehicle's Speed:
The c i e ion o speed in u ban en i onmen s o each ehicle a e Vmax and he minimum speed
is Vmin. We conside a uzzy se o { e y high, high, highe medium, medium, lowe medium,
low, e y low} o ehicle's speed. We conside he "medium speed" o speed c i e ion weigh .
 Fuzzy Va iable o he Numbe o Neighbo s:
In e na ional Jou nal in Founda ions o Compu e Science & Technology (IJFCST), Vol. 3, No.6, No embe 2013
67
We also conside a uzzy se including se en membe s { e y high, high, highe medium, medium,
lowe medium, low, e y low} o he numbe o neighbo s. Abou he c i e ion weigh o he
numbe o neighbo s, " he highes numbe o neighbo s “is conside ed.
 Fuzzy Va iable o Dis ance be ween Vehicles:
We conside a uzzy se o { e y high, high, highe medium, medium, lowe medium, low, e y
low} o he dis ance be ween ehicles. We use he equa ion in GPS o calcula e he c i e ion
weigh o op imal dis ance. We selec he weigh " e y high “ o dis ance c i e ion.
A e calcula ing he o al amoun s o c i e ia membe ship o e e y single ehicle, we s a o
calcula e uzzy decision-making in each ehicle acco ding o equa ions. A e calcula ion, we use
he g a i y cen e me hod o de uzzi ica ion o he ob ained amoun s. When b oadcas ing he
da a, each ehicle like Vi p io i izes he neighbo ing node acco ding o he upda ed da a o he
neighbo ing nodes and uzzy decision-making and selec s he node wi h high p io i y as he nex
s ep o send he da a. This me hod will conside ably dec ease he b oadcas ing a ic and will
inc ease i s e iciency.
The aim o he sugges ed me hod is o selec a node wi h highes numbe o neighbo s and
a e age speed and he possible a hes dis ance. The node ha ing he h ee c i e ion men ioned
abo e wi h high weigh s is selec ed as op imal node o da a b oadcas ing. I he link be ween he
p esen node and he node wi h high p io i y is b oken o some easons such as high speed, going
beyond 1000m sending limi and inc eased loading on ne wo k, hen a node wi h he nex p io i y
o de will be selec ed o da a b oadcas ing. The p io i y le el o each node will be c ea ed by
mul i-objec i e decision-making in a se acco ding o he o de p io i ies. In he neighbo ing able
o each node, he da a abou he neighbo ing nodes such as dis ance, numbe o neighbo s and
speed a e men ioned. Each node p io i izes hese nodes wi h uzzy decision-making. When a
ehicle wan s o b oadcas da a, a ehicle wi h highes p io i y is selec ed acco ding o he
a ailable p io i iza ion and ha ehicle will be used o ansmi he da a.
5. SIMULATION
Simula ion en i onmen is 4000×4000m and he maximum sending ange o each node is 1000m.
The packe s a e sen wi h ixed size o 128 by es and ixed a e o 4 pk s/sec. The maximum
numbe o packe s which can be sen in each sec ion is 6000 packs and can be ecei ed by 20
des ina ions. The calcula ion me hod o FMRBS in VANET was compa ed wi h o he mobile
models. The c i e ia e alua ed in simula ions include he ollowing cases:
Packe s dissemina ion Speed (m/s):is equal o he dis ance passed by a pack di ided by delay
Load Gene a ed pe B oadcas Packe : The numbe o bi s sen o b oadcas ing o a pack in he
whole ne wo k
The A e age Da a Packe Receip : exp esses he a e o ehicles ha ecei ed he b oadcas ed
pack o he o al packs
Me hods compa ed wi h ou sugges ed p o ocol include: UMB (U ban Mul i-Hop B oadcas
P o ocol) and 802.11-Dis ance.
In e na ional Jou nal in Founda ions o Compu e Science & Technology (IJFCST), Vol. 3, No.6, No embe 2013
68
5.1 Simula ion Resul s
- Packe s dissemina ion Speed:
In FMRBS, as packe s a e no sen o all neighbo s o b oadcas he da e and only he ehicles
wi h high p io i y a e selec ed as he nex s ep, he b oadcas ing speed o packs is mo e han ha
in o he me hods.
A e age packe dissemina ion speed (m/s)
A e age packe gene a ion a e o each ehicle (packe s/s)
Figu e 2. Packe s dissemina ion Speed
- Load Gene a ed pe B oadcas Packe
In b oadcas ing o da a using FMRBS me hod, ins ead o sending he pack o all neighbo s only
he neighbo ing ehicles wi h highe p io i y a e selec ed and he load on ne wo k in sending he
packs conside ably dec eases.
A e age load gene a ed o each b oadcas
packe (bi s)
A e age packe gene a ion a e o each ehicle (packe s/s)
Figu e 3. Load Gene a ed pe B oadcas Packe
- The A e age o Da a Pack Receip
As all ehicles a e no in ol ed in da a b oadcas ing, he ehicles wi h less neighbo s and low
speed canno ecei e a pack. The esul s o simula ion show ha his a e is e y low and is
almos equal o o he me hods.
In e na ional Jou nal in Founda ions o Compu e Science & Technology (IJFCST), Vol. 3, No.6, No embe 2013
69
A e age success pe cen age (%)
A e age packe gene a ion a e o each ehicle (packe s/s)
Figu e 4. The A e age o Da a Pack Receip
6.CONCLUSION
A new me hod based on Fuzzy Mul i-C i e ia Decision Making (MCDM) o ehicle-2- ehcile
ne wo ks, called FMRBS, o sol e he p oblem o B oadcas S o m in his pape . The nex node
o sending eme gency me hod is selec ed using uzzy ules in his me hod. The pa ame e s o
ehicle's speed, he numbe o neighbo s and dis ance be ween ehicles a e used as pa ame e s
e ec i e on de e mining he p io i y o ehicles o da a b oadcas ing. The ehicle wi h high
p io i y is used as he nex s ep in da a b oadcas ing.
In simula ion, he FMRBS me hod was compa ed wi h UMB and 802.11-dis ance me hods. The
esul s o simula ion show ha because o selec ing he mos p ope neighbo ing nodes as he nex
s age in da a b oadcas ing, he speed o sending he packs is inc eased and he ne wo k load is
conside ably dec eased. Also, because o high eliabili y o pa hs, sending he con ol packs is
conside ably inc eased.
REFERENCES
[1] Balon N., “Inc easing B oadcas Reliabili y in Vehicula Ad Hoc Ne wo ks,” In e na ional Wo kshop
on Vehicula Ad Hoc Ne wo k, pp. 104-105, 2004.
[2] Blum J., Eskanda ian A., Ho man L., “Challenges o In e ehicle Ad Hoc Ne wo ks,” IEEE
T ansac ion on In elligen T anspo a ion Sys ems, ol. 5, pp. 347–351, 2004.
[3] Fede al Communica ions Commission, “FCC 03-324. FCC Repo and O he ,” Feb ua y 2004.
[4] Jones W. D., “Keeping ca s om c ashing,” IEEE Spec um, olume. 38, pp. 40-45, Sep embe 2001.
[5] Kah aman CH., “Fuzzy mul i-c i e ia decision making,” Theo y and Applica ions wi h Recen
De elopmen s, Sp inge , 2008.
[6] Ka o S., Tsugawa S., Tokuda K., Ma sani T., Fujii H., “Vehicle con ol algo i hms o coope a i e
d i ing wi h au oma ed ehicles and in e ehicle communica ions,” IEEE T ansac ions on In elligen
T anspo a ion Sys ems, no. 3, pp. 155-16, 2002.
[7] Ku kmaz G., Ekici E., “U ban Mul i-Hop B oadcas P o ocol o In e -Vehicula Communica ion
Sys ems,” In e na ional Con e ence on Vehicula Ad hoc Ne wo ks, pp. 76-85, 2004.
[8] Kwang H. L., “Fi s Cou se on Fuzzy Theo y and Applica ions,” Sp inge , 2005.
[9] Ni S., Tsang Y., “The B oadcas S om p oblem in A Mobile Ad Hoc Ne wo k,” In e na ional
Con e ence on Mobile Compu ing and Ne wo king, pp. 11-162, 1993.
[10] T ian aphyllou E., “Mul i–C i e ia Decision Making Me hods: A Compa a i e S udy,” Kluwe
Academic Publishe s, 2000.
[11] Wang L.X., “a Cou se in Fuzzy Sys ems and Con ols,” P en ice-Hall In e na ional, Inc., 1997.