Show simple item record

dc.contributor.authorMaglaras, Leandrosen
dc.contributor.authorKatsaros, Dimitriosen
dc.date.accessioned2016-04-21T15:05:28Z
dc.date.available2016-04-21T15:05:28Z
dc.date.issued2015-01-20
dc.identifier.citationMaglaras, L. and Katsaros, D. (2016) Social Clustering of Vehicles Based on Semi-Markov Processes. IEEE Transactions on Vehicular Technology, 65 (1), pp. 318-322en
dc.identifier.issn0018-9545
dc.identifier.urihttp://hdl.handle.net/2086/11965
dc.descriptionThe full text version attached to this record is the authors final peer reviewed version. The publisher's final version of record can be found by following the DOI link.en
dc.description.abstractVehicle clustering is a crucial network managementtask for vehicular networks in order to address the broadcaststorm problem, and also to cope with rapidly changing networktopology. Developing algorithms that createstable clustersis avery challenging procedure because of the highly dynamic movingpatterns of vehicles and the dense topology. Previous approachesto vehicle clustering have been based on either topology-agnosticfeatures, such as vehicle IDs, on hard to set parameters, orhave exploited very limited knowledge of vehicle trajectories.This article develops a pair of algorithms, namelySociologicalPattern Clustering (SPC), andRoute Stability Clustering (RSC),the latter being a specialization of the former that exploit, forthe first time in the relevant literature, the “social behavior”of vehicles, i.e. their tendency to share the same/similar routes.Both methods exploit the historic trajectories of vehiclesgatheredby road-side units located in each subnetwork of a city, anduse the recently introduced clustering primitive ofvirtual forces.The mobility, i.e. mobile patterns of each vehicle are modeledas semi-Markov processes. In order to assess the performanceof the proposed clustering algorithms, we performed a detailedexperimentation by simulation to compare its behavior withthat of high-performance state-of-the-art algorithms, namely, theLow-Id,DDVCandMPBCprotocols. The comparison involvedthe investigation of the impact of a range of parameters onthe performance of the protocols, including vehicle speed andtransmission range as well as the existence and strength of socialpatterns, for both urban and highway-like environments. Allthe received results attested to the superiority of the proposedalgorithms for creating stable and meaningful clusters.en
dc.language.isoenen
dc.publisherIEEEen
dc.subjectClusteringen
dc.subjectmobilityen
dc.subjectsocial behavioren
dc.subjectMarkov processen
dc.subjectvehicular networksen
dc.titleSocial Clustering of Vehicles Based on Semi-Markov Processesen
dc.typeArticleen
dc.identifier.doihttp://dx.doi.org/10.1109/TVT.2015.2394367
dc.researchgroupCyber Security Centreen
dc.peerreviewedYesen
dc.funderEU.ICT program, ChallengeICT-2011.7en
dc.projectidREDUCTION: Reducing EnvironmentalFootprint based on Multi-Modal Fleet management System forEco-Routingand Driver Behaviour Adaptationen
dc.cclicenceCC-BY-NC-NDen
dc.date.acceptance2014-12-16en
dc.researchinstituteCyber Technology Institute (CTI)en


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record