EURASIPcentos6.5

RESEARCHOpenAccessOptimizationstrategyofmobiledatatransmissionbasedonoptimalcrowdfeedbackYangZhangAbstractInordertoeliminatethefactorsthatrestricttheperformanceofwirelessnetworkdatatransmission,weproposedtheoptimalcontrolmechanismofwirelessnetworkdatatransmission.
Theproposedmechanismsolvestheintelligenceproblemofthefeedbackloop,themobilityofrelaynodes,andthefeedbackofthereceivingend.
Ontheonehand,toeliminatetheexternalinterferencefactors,weestablishedanoptimalfeedbackloopcontrolsystembetweenthesenderandthereceiver.
Ontheotherhand,inthetimelinearregion,thecrowdfeedbackmoduleisaddedtotheoptimalfeedbackclosed-loopcontrolsystembasedonthelinearweight.
Onthebasisoftheaboveschemes,weproposedanadaptiveoptimizationmodelofmobiledatatransmission.
Theexperimentscomparedtheproposedoptimalcrowdfeedbackoptimalcontrolschemewiththeoptimizationstrategyofthedatatransmission.
Fromtheresultsofsystemefficiencyandsystemthroughputperformance,theproposedoptimalcrowdfeedbackoptimalcontrolschemehasanobviousadvantage.
Keywords:Controloptimization,Mobilitymanagement,Wirelessdatatransmission,Crowdfeedback1IntroductionThediversityof[1]anddataservicesinwirelesstrans-missionenvironment[2]makesthequalityofwirelesstransmissiondatadecrease.
Wirelesstransmissionsys-temperformanceisnotstable.
Theseproblemsser-iouslyaffectthequalityoftheuserexperience[3].
Theseissuesrestricttheway[4]andoperatingcostsofmobileservicecarriers.
Inordertosolvetheseprob-lems,theacademicandindustrialcircleshavestudiedthewirelesscommunicationprotocol[5–8],thewire-lessdatapacket[9,10],thetransmissionperformancecontroloptimization[11,12],andsoon.
JinMproposedaremotewirelesstransparenttrans-missionprotocolformeetingtherequirementsoftheInternetofThings-basedintelligentremotemonitoringsystemforconstructionvehiclesontherealtime[5].
Thedistance-sensitivewirelesscommunicationprotocolwasproposedinanarticle[6]tohearingprotectorsequippedwithin-earmicrophones.
ChaturvediAetal.
studiedthesecurewirelesscommunicationprotocolwiththeDiffie-Hellmankeyexchangescheme[7].
Thelinkperformanceofintra-satellitenetworkingwasanalyzedanddiscussed[8]forreceivingsensitivityineachmodel.
Thecellular/infostationintegratednetworkwascon-sideredbytheauthorsofanarticle[9]thatsupportson-demanddataservicedelivery.
Thenovelap-proacheswasproposedinanarticle[10]thatarebasedondataenvelopmentanalysis(DEA)tofurtheroptimizeenergyconsumptioninwirelessmulticastnetworks.
Therandomwirelesssensornetworkswereconsid-ered,wherenodesaredistributedrandomlyandformclusterstotransmitthepacketstorelayclustersusingcooperativemulti-input-multi-outputtechnique[11].
Asimpleandeffectivemethodisdemonstratedtoovercomethefrequencysplittingforanoptimalefficiency[12].
Therestofthepaperisorganizedasfollows.
Section2givestheoptimalfeedbackmodel.
Section3discussedtheoptimizationstrategyofmobiledatatransmissionwithcrowdfeedback.
ThealgorithmanalysisresultshaveCorrespondence:yanggzhang@yeah.
netCollegeofInternationalEconomicsandTrade,JilinUniversityofFinanceandEconomics,Jilin,ChinaEURASIPJournalonEmbeddedSystemsTheAuthor(s).
2016OpenAccessThisarticleisdistributedunderthetermsoftheCreativeCommonsAttribution4.
0InternationalLicense(http://creativecommons.
org/licenses/by/4.
0/),whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedyougiveappropriatecredittotheoriginalauthor(s)andthesource,providealinktotheCreativeCommonslicense,andindicateifchangesweremade.
ZhangEURASIPJournalonEmbeddedSystems(2016)2016:26DOI10.
1186/s13639-016-0062-6beeninSection4.
Finally,theSection5concludesthispaper.
2OptimalfeedbackmodelThelineardatatransmissionsystemforwirelesscom-municationnetworksisdescribedasfollows:yxHtxtFtdsytFxs1wherexsaidemissionsignalintensity.
tdenotesthedatatransmissiontime.
H(t)saidlinearvectordatatransmission.
FunctionF(t)saidlineardatatransmis-sioncontrolvector.
sisthelineartransmissiondatasequence.
d(s)denotesthesizeofthedata.
y(x)in-dicatesthetransmittingpowersignal.
y(t)representsthelineartimeonthetotaltransmitpower.
Wirelessdatatransmissionissubjecttovariousexter-naldisturbances,suchasobstaclesreflection,buildingdiffraction,groundabsorption,andotherfactors.
Anoptimalfeedbackloopisformedbetweenthesenderandthereceiver,whichisthekeytoeliminatetheinter-ferencefactors.
Theformula(2)describestheoptimalfeedbackclosed-loopcontrolsystems.
Fig.
1OptimalfeedbackloopSendingWirelesschannelMovingvectorReceivingfeedbackloopCrowdstackAdaptiveoptimizationCrowdfeedbackFig.
2AdaptiveoptimizationmodelofmobiledatatransmissionbasedoncrowdZhangEURASIPJournalonEmbeddedSystems(2016)2016:26Page2of5PtPsHtFxyt≥0Px≤dsuxtHs1FdXi1Nxsi8>>>>>>>:2Here,P(t)isthelineardataintimedomain.
N(x)rep-resentsthenumberofsymbolsinthedatasequencex.
uxtisthefeedbackloopofthedatasequencexinthetimedomain.
TheconstraintconditionP(t)+P(s)H(t)F(x)y(t)≥0canguaranteethestabilityofwirelessdatatransmission.
TheconstraintconditionP(x)≤d(s)canguaranteetheuniquenessandoptimalityoftheclosedloop(seeFig.
1).
3OptimizationstrategyofmobiledatatransmissionwithcrowdfeedbackTheoptimalfeedbackclosed-loopcontrolsystemiscomposedofmdatapackets.
krepresentsthenumberofpackets.
Thesystemtransmissionefficiencyηcanbeob-tainedfromtheformula(3).
ηXmk1dkds;wL0ηXmk1dkPx;wL>08>>>>>>>>>>>>>>>:3Accordingtothedifferenttimedomainlinearweight-ingwL,thesolutionisdividedintotwokinds.
Accordingtotheformula(2),itcanbeknownthattheoptimalsys-temefficiencycanbeobtainedwhenthewLisgreaterthan0.
InordertokeeptheprobabilityofwLbegreaterthan0,wedesignthecrowdfeedbackmoduleintheoptimalfeedbackloopcontrolsystem.
ThevaluefeedbackvectorxCRcanbeobtainedbytheformula(4).
xCRZt→TxtHxFtkkXmk1dkdkFdPx12Hs18>>>>>:4Thecrowdfeedbacksystemefficiencyofthetransmis-sionaftertheoptimization(seeFig.
2)canbecomputedwiththeformula(5).
ηCROPXmk1dkwLFxkkdxpPx5Afteroptimization,thetotallengthofthedatase-quenceLCR_OPcanbecalculatedbytheformula(6).
LCROPXmk1NxkwLXmk1dkFxkk264AlgorithmevaluationresultsThedatatransmittingendoftheexperimentalplatformisaclustercomposedoffiveservers.
ThereceiverrunsfiveLinuxvirtualsystemsinaparalleloperation.
Be-tweenthetransmitterandthereceiverisawirelesscom-municationnetworkcomposedof50mobilenodes.
Themappingrelationshipbetweenthetransmitterandtwovirtualsystemsofthereceiverisanendtoendmapping.
12534Wirelessnetworkwith50nodes12345ReceivingendSendingendMappingbetweenendsFeedbackFeedbackFig.
3ExperimentalarchitectureTable1ExperimentalsettingsParametersValueParametersValueCPU8coreMemory32GBOSCentOS6.
5Harddiskspacesize1TBDriverFaban-kitDatabaseSQLServerMonitorVmstatNumberofconcurrentsession1,2,3ZhangEURASIPJournalonEmbeddedSystems(2016)2016:26Page3of5Therelationisusedtoestablishtheoptimalcrowdfeed-backclosed-loopcontrolsystem.
TheexperimentalstructureisshowninFig.
3.
TheserverhardwareandsystemsoftwareconfigurationisshowninTable1.
Intheexperiments,wecomparedtheproposeddatatransmissionoptimization-optimalcrowdfeedback(DTO-OCD)withthedatatransmissionoptimizationstrategybasedontheoptimizedfeedbackloopdatatransmissionoptimization-feedbackloop(DTO-FL)withthesystemefficiencyandthroughputperformance.
Figures4,5,and6givetheDTO-OCDandDTO-FLsystemefficiencywiththechangeofthedatapacketsizewhenthepacketlossrateis0.
5,1,and5%,respectively.
Foundthatwiththeincreaseofthepacketlossrate,thesystemefficiencyofthetwomechanismsbegantosignificantlyreduce.
Withtheincreaseofthedatapacket,thesystemefficiencyofthetwomechanismsischattering.
However,thesystemefficiencyofDTO-FLdecreasedmoreobviously.
ItsdeclinerangeisthesameortwotimesasmuchasthatofDTO-OCD.
DTO-FLjitterismoreefficientthanthatofDTO-OCD.
Figures7,8,and9givethesystemthroughputratere-sultswiththechangeofthedatapacketsizewhenthewirelessnetworkmobilespeedis0.
5,1,and5m/s,re-spectively.
Withtheincreaseofthespeedofnodemove-ment,wefoundthattheexternaldisturbanceisthekeyfactortocontrolthequalityofdatatransmission.
Thesystemthroughputrateofthetwomechanismsbegantosignificantlyshake.
Thejitterisespeciallyevidentwhenthemovingspeedis5m/s.
Inaddition,withtheincreaseofthedatapackage,thesystemthroughputdeclined.
However,thesystemefficiencyofDTO-FLdecreasedmoreobviously.
ThedeclinerangeofDTO-FLis2timestheoneofDTO-OCD.
Fig.
4Systemefficiencywhenpacketerrorrateis0.
5%Fig.
5Systemefficiencywhenpacketerrorrateis1%Fig.
6Systemefficiencywhenpacketerrorrateis5%Fig.
7Throughputwhenmovingspeedis0.
5m/sZhangEURASIPJournalonEmbeddedSystems(2016)2016:26Page4of55ConclusionsBasedonthestudyofthedeclinereasonsofthedatatransmissionperformanceinthewirelessnetworkenvir-onmentandtheinfluencingfactorsofthefeedbackmechanism,weputforwardtheoptimalcrowdcontrolofthewirelessnetworkdatatransmission.
Theinnovationoftheschemehasalsoacrowdfeedbackmechanism,dataencapsulation,mobilemanagement,andsensoryfeedback.
Firstly,aclosed-loopcontrolsys-temisformedbetweenthetransmitterandthereceiver,whichisaffectedbytheinterferenceofthewirelessdatatransmission.
Theaimofthesystemistoeliminatethekeyfactorsofthesefactors.
Secondly,accordingtothetimedomainlinearweight,crowdfeedbackmoduleisdesignedtotheoptimalfeedbackclosed-loopcontrolsystem.
Finally,onthebasisoftheaboveresearches,weproposedanadaptiveoptimizationmodelofmobiledatatransmission.
Basedontheexperimentalresults,wefoundthatthesystemefficiencyandthesystemthrough-putrateoftheproposedalgorithmarehigherandthesystemthroughputishigherandsmoother.
CompetinginterestsTheauthordeclaresthathe/shehasnocompetinginterests.
Received:18August2016Accepted:11November2016References1.
CerpaA,ElsonJ,EstrinD,etal.
Habitatmonitoring:applicationdriverforwirelesscommunicationstechnology[J].
AcmSigcommWorkshoponDataCommunicationsinLatinAmerica&theCaribbean,2010;31(2):20–41.
2.
ChamotraS,SehgalRK,KamalR,etal.
Datadiversityofadistributedhoneynetbasedmalwarecollectionsystem[C]//EmergingTrendsinNetworksandComputerCommunications(ETNCC),2011InternationalConferenceon.
IEEE,2011:125–129.
3.
DMLestari,DHardianto,HANizar,Analysisofuserexperiencequalityonresponsivewebdesignfromitsinformativeperspective[J].
Int.
J.
Softw.
Eng.
ItsAppl.
8(5),53–62(2014)4.
LDeng,QZeng,WZhouetal.
,Theeffectoftrainformationlengthandservicefrequencyonthedeterminationoftrainschedules[J].
ProceedingsoftheInstitutionofMechanicalEngineersPartF.
J.
RailRapidTransit.
228(4),378–388(2013)5.
MJin,XZhou,ELuoetal.
,Industrial-QoS-orientedremotewirelesscommunicationprotocolfortheinternetofconstructionvehicles[J].
IEEETrans.
Ind.
Electron.
62(11),7103–7113(2015)6.
REBSerhal,THFalk,JVoix,Integrationofadistancesensitivewirelesscommunicationprotocoltohearingprotectorsequippedwithin-earmicrophones.
[J].
J.
Acoust.
Soc.
Am.
133(5),3273(2013)7.
AChaturvedi,NSrivastava,VShukla,AsecurewirelesscommunicationprotocolusingDiffie-Hellmankeyexchange[J].
Int.
J.
Comput.
Appl.
126,33–36(2015)8.
QQi,PLi,XUGuo-Dong,Designandsimulationofthewirelesscommunicationprotocolappliedtosmallsatellite[J].
LEMCS-1438(1),71–76(2014)9.
TChen,HShan,XWang,Optimalschedulingforwirelesson-demanddatapacketdeliverytohigh-speedtrains[J].
IEEETrans.
Veh.
Technol.
64(9),4101–4112(2015)10.
AAAjibesin,NVentura,AMurguetal.
,Dataenvelopmentanalysiswithslacksmodelforenergyefficientmulticastovercodedpacketwirelessnetworks[J].
IetSci.
Meas.
Technol.
8(6),408–419(2014)11.
BLi,HLi,WWang,QYin,HLiu,Performanceanalysisandoptimizationforenergy-efficientcooperativetransmissioninrandomwirelesssensornetwork[J].
IEEETrans.
Wirel.
Commun.
12(9),4647–4657(2013)12.
JHan,YDKim,NHMyung,Efficientperformanceoptimisationofwirelesspowertransmissionusinggeneticalgorithm[J].
Electron.
Lett.
50(6),462–464(2014)Submityourmanuscripttoajournalandbenetfrom:7Convenientonlinesubmission7Rigorouspeerreview7Immediatepublicationonacceptance7Openaccess:articlesfreelyavailableonline7Highvisibilitywithintheeld7RetainingthecopyrighttoyourarticleSubmityournextmanuscriptat7springeropen.
comFig.
8Throughputwhenmovingspeedis1m/sFig.
9Throughputwhenmovingspeedis5m/sZhangEURASIPJournalonEmbeddedSystems(2016)2016:26Page5of5