An Efficient Neural Network-Based Prediction Scheme for Heterogeneous Networks

Inmobilewirelessnetworks,thechallengeofprovidingfullmobilitywithoutaffectingthequalityof service(QoS)isbecomingessential.Thesechallengescanbeovercomeusinghandoverprediction. Theprocessofdeterminingthenextstationwhichmobileuserdesirestotransferitsdataconnection canbetermedashandoverprediction.Anewproposedpredictionschemeispresentedinthisarticle dependentonscanningallsignalqualitybetweenthemobileuserandallneighboringstationsinthe surroundingareas.Additionally,theproposedschemeefficiencyisenhancedessentiallyforminimizing theredundanthandover(unnecessaryhandovers)numbers.BothWLANandlongtermevolution (LTE)networksareusedintheproposedschemewhichisevaluatedusingvariousscenarioswith severalnumbersandlocationsofmobileusersandwithdifferentnumbersandlocationsofWLAN accesspointandLTEbasestation,allrandomly.Theproposedpredictionschemeachievesasuccess rateofupto99%inseveralscenariosconsistentwithLTE-WLANarchitecture.Tounderstandthe networkcharacteristicsforenhancingefficiencyandincreasingthehandoversuccessfulpercentage especiallywithmobilestationhighspeeds,aneuralnetworkmodelisused.Usingthetrainednetwork, itcanpredictthenexttargetstationforheterogeneousnetworkhandoverpoints.Theproposedneural network-basedschemeaddedasignificantimprovementintheaccuracyratiocomparedtotheexisting schemesusingonlythereceivedsignalstrength(RSS)asaparameterinpredictingthenextstation. Itachievesaremarkableimprovementinsuccessfulpercentageratioupto5%comparedwithusing onlyRSS.