简介：Greyself-organizingmap(GSOM)modelisproposedandappliedinthedetectionofintrusion.ThroughtheimprovementoftheweightadjustmentusingtheGRC(greyrelationalcoefficient),thetrainingresultsofSOMgetbetter.Inthedetectionofdenyofservice(DOS)attacks,thismodelcanconsidertherelativityofthedatasetofDOSattacks.Finally,theexperimentsontheDOSdatasetconfirmtheirvaliditiesandfeasibilitiesoverthisGSOMmodel.

简介：Apeer-to-peer(P2P)networkisadistributedapplicationarchitecturewhichprovidesmanyattractivefeatures,suchasavailability,self-organization,load-balancing,andanonymity.However,P2Pnetworkhascreatedsignificantproblemstonetworkoperatorsbygeneratinglargevolumesofinterautonomoussystem(inter-AS)traffic.FocusingontheBitTorrentswarmingprotocol,thispaperproposesanapproachwhichaimstoreduceP2Pgeneratedinter-AStraffic.Inparticular,theapproachcanreduceinter-AStrafficby50%to70%.Moreover,itcanimprovethedownloadingspeedby60%forthepopulartorrents.Theevaluationshowsthatcontrolledregional-basedcontentsreplicationcaneffectivelyachievethisgoal.Furthermore,theapproachisincrementallydeployable.NetworkregionsinwhichthesystemgetsdeployedcansolvetheirP2Pgeneratedinter-AStrafficproblemsautonomously,i.e.,withoutanyInternetserviceproviders-collaborationandanyrequirement,thesystemcanbedeployedintheentireInternet.

简介：Low-duty-cyclemechanismscanreducetheenergyconsumptionsignificantlyinwirelesssensornetworks(WSNs).Sensorsstaydormantmostofthetimetosavetheirenergyandwakeupbasedontheirneeds.However,suchatechnique,whileprolongingthenetworklifetime,setsexcessivechallengesforreducingtheend-to-end(E2E)delaywithinthenetwork.Inthispaper,thecentralizedcluster-basedlocationfinding(CCLF)algorithmisproposedtoreducethehighlatencyinlow-duty-cycleWSNsbyfindingasuitablepositionforthesink.Thealgorithmismainlycomposedofthreesteps:a)theclusterconstruction,b)thefastlook-uptable(FLU-table)construction,andc)thesinklocationdecision.ThesimulationresultsshowthattheperformanceoftheCCLFalgorithmissignificantlysimilartothatoftheoptimalalgorithm.Moreover,theCCLFalgorithmrequireslessoperationtimecomparedwiththeoptimalalgorithm.

简介：Reversibledatahidingtechniquesarecapableofreconstructingtheoriginalcoverimagefromstego-images.Recently,manyresearchershavefocusedonreversibledatahidingtoprotectintellectualpropertyrights.Inthispaper,wecombinereversibledatahidingwiththechaoticHénonmapasanencryptiontechniquetoachieveanacceptablelevelofconfidentialityincloudcomputingenvironments.And,Haardigitalwavelettransformation(HDWT)isalsoappliedtoconvertanimagefromaspatialdomainintoafrequencydomain.Andthenthedecimalofcoefficientsandintegerofhighfrequencybandaremodifiedforhidingsecretbits.Finally,themodifiedcoefficientsareinverselytransformedtostego-images.

简介：