Journal iof iScience iand iTechnology i

ISSN: i2456-5660 iVolume i4, iIssue i5, iSep-October i2019, iPP i01-06
www.jst.org.in
ii Performance Analysis of Cognitive Radio Network Using Enhanced
i i i i i i i

Stimulated Annealing Algorithm i i i

N V S S Ramya Jyothi1, Dr.B. LeelaKumari2

i i i i i ii i
1
(M.Tech iScholar, iECE iDepartment, iJNT iUniversity, iKakinada) i
2
(Assistant iProfessor, iECE iDepartment, iJNT iUniversity, iKakinada) i
________________________________________________________________________
Abstract: iThis idocument ipresents ithe ioptimization iof iquality iof iservice ifor iunlicensed isubscribers ianticipating ithe iQoS iof
ithe iprimary iusers ias iwell. iThe iconcept iof icognitive iradio icame iinto iexistence ifor ithe iexclusive iuse iof ithe iunused ispectrum
ispace iby ithe ilicensed iusers. iThe iunlicensed iuser ior ithe isecondary iusers iuse ithe iunoccupied ispectrum ispaces iof ilicensed
iuser's ispectrum ifor ithe ieffective iusage iof ithe ispectrum. iThe iuse iof ilive istreaming, iVoice iover iInternet iprotocol(VoIP) iand

imulti imedia iapplications iwhich iare idelay isensitive iapplications ithroughout ithe isession. iThe iconstraints ilike idelay iand

ithroughput iand ithe iquality iof iservice iare iaffected idue ito ithese iapplications. iTo iovercome ithese iconstraints ithe itradeoff

ibetween ithese iparameters imust ibe iemployed. iIn ithis ipaper, ian iextension ito ithe inon-work iconservation ipolicy iis iimplemented

iin iCognitive iradio inetwork(CRN) iby iusing ithe ioptimization ialgorithm iknown ias iEnhanced istimulated iannealing(ESA)

ialgorithm ito ibring iforth ithe itradeoff ibetween idelay iand ithroughput.

Keywords: iCognitive iRadio, iDelay, iEnhanced istimulated iannealing iAlgorithm, iOptimization, iThroughput, iTrade-off.
___________________________________________________________________________________________________

I. Introduction
With ithe igrowth iin ithe idigital iage, inowadays ispectrum iusage ihas irisen idrastically. iThe iuse iof ispectrum ihas ithe imost
iprominent iposition ito iplay iin ithe iworld iof icommunication. iThe itechnology icalled icognitive iradio ihas ideveloped iby ithe iIEEE

i802 iLAN i/ iMAN iStandard iCommittee i(LMSC) iand ireleased iin i2011. iIt iis ian iintelligent iradio ithat ican ichange ithe iparameters

iin iaccordance ito ithe iperceived iavailability iof ithe ispectrum iin iits ioperating ienvironment. iCognitiveradio isupports ithe

icognitive iusers ialso iknown ias ithe isecondary iusers ior ithe iunlicensed iuser iand ican iprovide ithem iunused ispectrum ispace ito

ieffectively iuse ithe iunderutilized iband ifrequencies. iA icognitive iradio i(CR) ihas ithe icapacity i(cognitive icapacity), ifrom ithe

inearby ienvironmentto idetect iand icollect idata i(e.g., itransmission irate, ibandwidth, ienergy, imodulation, ietc.) ias iwell ias ito

iadjust ioperating iparameters ito iideal iresults iquickly i[1]. iThe icognitive icycle ihas ithree iphases. iThey iare i(i).spectrum isensing,

i(ii). ispectrum imanagement, i(iii). ispectrum imobility, i(iv). ispectrum isharing. iThe icognitive icycle iis ias ishown iin ithe ifollowing

ifigure i1.

Figure i1: iCognitive icycle

iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii www.jst.org.in i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i1 i| iPage

Journal iof iScience iand iTechnology i
ISSN: i2456-5660 iVolume i4, iIssue i5, iSep-October i2019, iPP i01-06
www.jst.org.in
The ispectrum isensing iis ithe iprocess ito isense ithe ispectrum iholes iin ithe ifrequency irange iof ithe isystem iin iwhich ithe
icognitive iradio iis iimplemented. iSpectrum isensing idetects ispectral iholes iand iallocates ifor ithe iopportunistic itransmission iof
iCR iusers. i iIn i[3] iMultiple isensors ifor iunified ispectrum idetection, iwhile ia imulti-energy idetector isystem iand ivarious

iadjustable ithresholds ifor icooperative ispectrum imeasurement iwere idescribed iin i[4]. iFurthermore,the ispectrum imanagement

iis ito ichoose i

the ibest ichannelswhich iare iavailable iand iarrange ithe iconnectivity iwith iother iusers ito ithis ichannel. iThe ispectrum
imanagement iis iall iabout iavoiding ithe iinterference iand iproviding iQoS iin ispectrum ienvironment. iAlong iwith iprovisioning iof
iseamless iconnectivity inevertheless iof ithe ilicensed isubscribers. iThe ispectrum imobility iis iregarding ithe ihandoff iduring ithe

ilicensed iuser iactivity ion ithe iallotted ichannel imight irequire iusers ito ialter itheir iworking ichannel ibands. i

A isignificant icondition iof imobility imanagement iprotocols iis ito iprovide idata ion ithe ispan iof ia ispectrum ihandover.
iWhenever ithe icognitive iuser ichanges iits ioperating ifrequency ithese inetwork iprotocols ichanges ithe iparameters. iThis ileads ito
ithe iguaranteed iclean iand irapid ishift iwith ia iminimum idepletion iof ithe iQoS iduring ihandovers.The icurrent ispectrum isharing

iresearch iis iintended iat iaddressing ithese iproblems iand ican ibe icategorized iinto ifour ielements: idesign, iconduct iof ispectrum

iallocation, imethod iof ispectrum icommunication, iand irange. iThe ifollowing isketch iexplains ithe iinter-network iand iintra

inetwork ispectrum isharing ii.e., ifigure i2.

Figure i2: iSpectrum isharing

II. Organization iOf iThe iFrame iWork

In ithis ipaper iwe ihave iconsidered ithe icognitive iradio inetwork ias isketched ias ibelow ifigure i3. iThe inetwork iincorporates iof
itwo iusers ii.e., iPrimaryusers i(PU) iand isecondary iusers i(SU) iand ia istandard idestination. iThe iPU iis ifurnished iwith ia iqueue,

iQp ifor ithe iprimary iuser ipackets iand ilikewise iSU iis ifurnished iwith ia iqueue, iQs imeantfor ithe isecondary iuser ipackets iand ithe

iQsp iwhich iis iintended ifor ithe ipackets i

iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii www.jst.org.in i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i2 i| iPage

 Journal iof iScience iand iTechnology i
ISSN: i2456-5660 iVolume i4, iIssue i5, iSep-October i2019, iPP i01-06
www.jst.org.in
Figure i3: iSystem iorganization

over iheaded ifrom ithe iPU ipackets. iThe ipacketarrival irates iare idefined ias iλp iand iλs iusing irandom iprocess iof iBernouliwhere
i0 i≤ iλp i≤ i1 iand i0≤ iλp i≤ i1.The ievolution iof ithe ilength iof ithe ijth iqueue iischaracterized iasQjt+1 i= i(Qtj i– iYtj) i+ i+ iXtj, ifor ij i∈

i{p, isp, is} i……. i(1)When ia iprimary iuser iuses ithe ichannel ito ipropagatethe ipackets ithrough iQp itowards ithe idestination,

ithepackets iare ireceived iby ithe idestination ia ipositiveacknowledgment i(ACK) iis isend ito iPU iindicating ithatthe ipackets iare

ireceived isuccessfully. iIn ithis iscenariowe iassume ithat ithe isystem iis iinterference ifree ichannelby iusing ithe ihigh idetection

iprobability idetectors iat itheMAC ilayer. iThe iprobability iof isuccessful ipacketreception ii.e., ithe iprobability iof ino ilinkage

ioutage,between ithe iprimary iuser i(PU) iand ithe idestination, ithesecondary iuser i(SU) iand ithe idestination, iand ithe iPUand ithe

iSU iare iindexed iby ihpd, ihsd iand ihps,respectively. iIn ithis ilayer ithe iPU ican ibe idetected ifor iitsIdle ior ithe ibusy istates iusing ithe

idetecting isensors iatthe isecondary iusers.

2.1 iWhen ithe iPU iis ireserved

Basically, iwhen ithe iPU itransmittes ithe ipacketstowards ithe idestination, ithere iexists ithree iprobablecases:

(i). iIf ithe idestination ireceives ithe ipackets iand isendsback ian iacknowledgement ito ithe iPU ithat ithe ipacketsare idecoded
iperfectly iby ithe idestination ineverthelessof iwhether ithe iSU idecoded ithe ipackets ior inot.

(ii). iIf ithe ipackets iare idecoded iby ithe iSU ias iwell ias ithedestination, ithen ithe ipackets iare istored iin iQSP iwithprobability
idenoted iby i‘a’. iIf iaccepted iby ithe iSU ithen iitacknowledges ithe isource iand ithe ipackets iare ireleasedfrom ithe iQp.

(iii). iIf ineither ithe iSU inor ithe idestination, ithen ithepackets iare iremained iin ithe iQp ifor iretransmission.

2.2. iWhen iPU iis inot ibusy

If ithe ichannel iis ioccupied iby ithe iSU ithrough iQs iwiththe iprobability i‘b’ ior ifrom ithe iqueue iQsp iwith iprobability‘1-b’.
i Otherwise ithe ipackets iare idropped, iand ithe iPUis iidle. iThe icondition ithat ithe iPU iis ibusy, iand ithe islotis iunder iutilized iis isaid
i to ibe ithe iWC icooperative ipolicy.

III. Implementation iOf iDesign

The iservice irates iof ithe iPU iand iSU iare iformulated ias ibelow ifrom ithe iservice irate iof iQp iis iμp, i

μp i= ihpd i+ i(1 i− ihpd) ihps ia

and isimilarly, ithe iservice irate iof iQs iis iμs, i

μs i= ib ihsd(1 i– iλp/μp)

whereas ithe iservice irate iof iQsp iis iformulated ias

μsp i= i(1 i− ib) ihsd i(1 i- iλp/μp).

The iaverage iqueue ilengths iof ithe iQS, iQp iand iQsp iare iindicated iby iNp, iNsp iand iNs iindividually. iThese iqueue ilengths iare
iderived ifrom ithe iapplication iof iPollaczek-Khinchine iformula i[8] iwith iBernoulli iarrival irate iwhich iis irepresented iby iλp iusing

idiscrete itime iM/M/1 iqueue iin ithe iMATLAB iusing icommunication itools.

The iInequalities iof ithe istability iof ithe iqueue iis iformulated isuch ithat ithe inetwork iwould ibe ioperated iaccordingly. iThat iis

λp i< iμp, iλs i<μs, iλsp<μsp.

Likewise, ithe iaverage idelays iof ithe iprimary iand icognitive iuser iare iformulated iusing ilittle’s ilaw isuch ithat

Dp i= i(Np i+ iNsp)/λp, iDs i= iNs/λs

iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii www.jst.org.in i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i3 i| iPage

Journal iof iScience iand iTechnology i
ISSN: i2456-5660 iVolume i4, iIssue i5, iSep-October i2019, iPP i01-06
www.jst.org.in
For idefining ithe isecondary iuser ithroughput, ithe ioptimization iproblem iis iconsidered ias ifollows. iHere iwe ioptimized ithe
iproblem iof imaximizing ithe iSU ithroughput iwith ithe iPU idelay iconstraint. iHere iis ithe iformula ithat ithe iproblem iis iframed:

Q1: imax ibhsd i(1- i(λp/ iμp))

a,b

s.t. i0 i≤ ia i≤ i1,

iiiiiiiiiiiiiiiiiiiiiiiiiiii 0 i≤ ib i≤ i1,

μp i= ih ipd i+ i(1 i− ih ipd) ihps ia,

(Np i+ iNsp)/λp i≤ iψ i[8]

The ipaper iis imainly ifocused ion ithe iconcept ito iimprove ithe iQoS ieven iwith icognitive iuser iwhile iusing ithe
iapplications ilike iVoice iover iinternet iprotocol, ilive istreaming iand isome iother imedia iapplications ilike iscreen isharing ietc.
iParallelly iproviding ithe ilicensed iuser iwith ithe iefficient icommunication ichannel.

The isuboptimal ipolicy iand ithe iWC ipolicy iare iprincipled iwith ithe iconditions isuch ias iarrival irates iand ithe
iprobabilities iof ithe ichannel ias ifollows: ih ipd i= i0.3, ih ips i= i0.4, ihsd i= i0.8. iNow icoming ito ithe iPrimary idelay iconstraint, ithe
iThreshold ilimit iplays ia iimportant irole iwhere iψ i= i10. iThe iproblem iQ1 iis ia inon-convex ifunction iwhich iis iconverted ito iquasi

iconvex ifunction ito isolve iin ian iiterative iprocess iwith iψ i= i10 iand iλp i= i0.2 iand iλs i=0.4.

IV. Stimulated iAnnealing iAnd iIts iAdvantages

It iis ia iprocess ipurely ibased ion ithe istaticalmechanics iand iAnnealing iis ia iprocess iin iwhich ithe imetallic isubstance iis iheated
iup iand icooled ito imake iit istronger ithan ibefore. iThe istimulated iannealing ialgorithm ihere iused ito isolve ischeduling iprocess iin
ithe icognitive iradio ito iavoid idelay iand ito iimprove ithe iperformance iof ithe iservice. iIt iis ivery iclear ithat ithe istimulated

iannealing iprocess iis ithe ibest ianswer ifor ithe ioptimization ithan ithe iconventional ialgorithms iused. iEnhanced istimulated

iannealing iprocess iinvolves ithe ifollowing isteps:

(i) The itasks ion ithe isystem iare idistributed iArbitrarily iand iassess ithe iweight iof ithe ischeme i(using ithe iOFso) ifor i‘n’ inumber
iof itimes.

(ii) iFind ithe iaverage iweight iof in-unit iweights idiscovered iin iportion i1.
(iii) To iconduct imulti-thread iannealing, iselect ia ilimit iproportion iso ithat iany iinstance iwith ia iunit iweight ihigher ithan i1/5
itimes iof i‘n’ ivalue i(provided) iabove ithe imedian iis ichosen ias ia iseed.

(iv) Collect ias imany iseeds ias ipossible.

(v) Apply ifor ieach iseed ithe isimulated iannealing imethod.
(vi) Select ithe iallocation iwith ithe iminimum iweight iof ithe ischeme.
Here ithe itemperature iis ia iparameter iused ito iindicate ithe iacceptance iratio iin ithe igiven isystem iand iinitial iparameter
ivalue iis ifixed isuch ithat ithe iproportion iof ithe iworst imoves iof igiven isystem iis iminimum.We isuggest ia ifresh isimulated

iannealing imethod idepending ion ithe iconcept iof iestimating ieach ichain's itemperature idepending ion ithe ilikelihood iof

iaccepting ithe igoal. iThe ievolution iof iacceptance iprobability iduring ioptimization ifollows ian iacceptance ischedule: iinitial

ivalue, ichain ilength, ifinal ivalue i(or istopping icondition) iand idecrement irule.The isystem iweight ienhancement ifor

iEnhanced istimulated iannealing iprocess iis imuch igreater ithan ithe i simple istimulated iannealing iprocedure ias ifollow iin

ithe ibelow iequation iµ i[Enhancement(Enh_SA)]>µ i[Enhancement(SA)] i. iSA iis isensitive ito ithe ivariation iof ithe icooling

ifactor.When ithe icooling ifactor iincreases, ithe ienhancement ifrom ithe iinitial isystem iweight iand ithe iexecution itime ialso

iincrease.

iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii www.jst.org.in i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i4 i| iPage

Journal iof iScience iand iTechnology i
ISSN: i2456-5660 iVolume i4, iIssue i5, iSep-October i2019, iPP i01-06
www.jst.org.in

Figure i4: iflowchart ifor iEnhanced istimulated iannealing(ESA)

In iaddition, ithe ienhancement ifrom ithe iinitial isystem iweight iand ithe iexecution itime iwere ifound ito iincrease iwith ithe
i increase iof ithe icooling ifactor ifor ithe isimulated iannealing ialgorithm. i

V. Summary iof iResults

The ifollowing iare ithe igraphical irepresentations iof ithe iaverage idelay ipacket iversus iarrival irate iat iSU ipackets, ithe istable
i throughput iregion igraphs iand ithe igraph ithat iillustrates ithe iaverage ipacket idelay iof ithe ilicensed iand iunlicensed iusers.

Figure i5: iGraph ithat iillustrating ithe iDelay iof iPU iVs iarrival irate ion iPU iwith ipredefined iconditions iand ithreshold.

Figure i6: iGraph iillustrating ithe iaverage iPU ipacket idelay iVs iaverage iSU ipacket idelays icomparing iwith ithe iimplementation
iof iESA.

iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii www.jst.org.in i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i5 i| iPage

Journal iof iScience iand iTechnology i
ISSN: i2456-5660 iVolume i4, iIssue i5, iSep-October i2019, iPP i01-06
www.jst.org.in

Figure i7: iGraph ithat iillustrates ithe iarrival irate iand ithe idelay ion iSU iwith ipredefined iconditions iand ithreshold.

Figure i8: iGraph iillustrating ithe iDelay iand ithroughput itrade ioff iat ithe iSU iconsidering ilicensed iusers iDelay iconstraint iwith ia
iψ=10.

VI. Conclusion
The iimplementation iof ithe iESA iover ithe iconventional iwork iconservative ialgorithm iand ithe inon-work iconservative ialgorithm
iflowered iin ibest itradeoff ibetween ithe icontroversial iparameters. iThe itradeoff ior itheimprovementis iabout i10%.Theresult iis

ifinalized ibased ion ithe ithreshold iparameter iat idifferent ivalue iof iaverage ipacket idelay ithat ithe ilicensed isubscribers ican

iwithstand. iThe iparameters isuch ias iarrival irates iand ithe iprobabilities iof ithe ispectrum iused iare iasfollows:

h ipd i= i0.3, ihps i= i0.4, ihsd i= i0.8

References
[1] iHaykin, iSimon. i(2005). iHaykin, iS.: iCognitive iradio: iBrain-empowered iwireless icommunications. iIEEE iJSAC i23(2), i201-220. iSelected iAreas iin
iCommunications, iIEEE iJournal ion. i23. i201 i- i220. i10.1109/JSAC.2004.839380.
[2] iAdel iM. iElmahdy, iAmr iEl-Keyi, iTamer iElBatt, iSenior iMember, iIEEE, iand iKarim iG. iSeddik, iSenior iMember, i“Optimizing iCooperative iCognitive
iRadio iNetworks iPerformance iwith iPrimary iQoS iProvisioning”, iIEEE iTRANSACTIONS iON iCOMMUNICATIONS, iVOL. i65, iNO. i4, iAPRIL i2017.
[3] iJ.kanti, iG. iS. iTomar, iand iA. iBagwari, i“A inovel imultiple iantennas ibased icentralized ispectrum isensing itechnique,” iin iTransactions ion iComputational
iScience, ivol. i10220 iof iLecture iNotes iin iComputer iScience, ipp. i64–85, iSpringer, iBerlin, i2017.
[4] iA. iBagwari, iG. iS. iTomar, iand iS. iVerma, i“Cooperative ispectrum isensing ibased ion itwo-stage idetectors iwith imultiple ienergy idetectors iand iadaptive
idouble ithreshold iin icognitive iradio inetworks,” iCanadian iJournal iof iElectrical iand iComputer iEngineering, ivol. i36, ino. i4, iArticle iID i6776586, ipp. i172–
180, i2013.
[5] iL. iKleinrock, iQueueing iSystems: iTheory, ivol. i1. iNew iYork, iNY, iUSA: iWiley, i1975.
[6] iBen-Ameur, iWalid.(2004). iComputing ithe iInitial iTemperature iof iSimulated iAnnealing. iComputational iOptimization iand iApplications.29.369-
385.10.1023/B:COAP.0000044187.23143.bd.
[7] iG.E. iNasr, iA. iHarb iand iG. iMeghabghab
“Enhanced iSimulated iAnnealing iTechniques ifor iMultiprocessorScheduling” iProceedings iof ithe i12th iInternational iFLAIRS iConference.
[8] iMohdRizam iAbu iBakar, iAbdul iJabbar iKhudhurBakheet, iFarah iKamil, iBayda iAtiya iKalaf, iIraq iT. iAbbas, iand iLee iLai iSoon, i“Enhanced iSimulated
iAnnealing ifor iSolving iAggregate iProduction iPlanning,” iMathematical iProblems iin iEngineering, ivol. i2016, iArticle iID i1679315.
[9] iM. iAshour, iA. iA. iEl-Sherif, iT. iElBatt, iand iA. iMohamed, i“Cognitiveradio inetworks iwith iprobabilistic irelaying: iStable ithroughput iand
idelaytradeoffs,” iIEEE iTrans. iCommun., ivol. i63, ino. i11, ipp. i4002–4014,Nov. i2015.
[10]A. iM. iElmahdy, iA. iEl-Keyi, iT. iElBatt, iand iK. iG. iSeddik, i“On ioptimizingcooperative icognitive iuser iperformance iunder iprimary iQoSconstraints,” iin
iProc. iIEEE iWireless iCommun. iNetw. iConf. i(WCNC),Apr. i2016, ipp. i1–7.

iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii www.jst.org.in i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i6 i| iPage