Font Size: a A A

Research On Data Aggregation In Vehicular Ad Hoc Networks

Posted on:2017-01-13Degree:MasterType:Thesis
Country:ChinaCandidate:S N WengFull Text:PDF
GTID:2322330512472022Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
Vehicular Ad Hoc Network(VANET)is an important part in intelligent transportation scheduling.In VANET,due to vehicular safety,traffic monitoring,and other commercial.applications,each vehicular node needs to broadcast information periodically,it could cause conflict and congestion in network,and consume a lot of bandwidth.Therefore,reduce network performance.In addition,the quick movement of vehicle will cause link breakage,improve network delay and the cost of network control.Therefore,an efficient data aggregation approach to aggregate data with less bandwidth consumption and high accuracy under a highly dynamic network is designed for reducing communication and maintaining cost.Considering the influence in network structure and studying the key technology of clustering and intra-cluster data aggregation,we focus on dynamic clustering algorithm and intra-cluster data aggregation methods,to solve the problem of quick change network in aggregation dynamic network structure.Firstly,we propose a clustering algorithm based on dominated set,which is short for dominating set based cluster algorithm(DSCA).In order to reduce the cost of cluster head selection and cluster maintaining,we propose a clustering algorithm based on dominated set.DSCA selects cluster head based on leader factor and relative movement,and build cluster by optimistic leader.With the change of cluster topology,in order to reduce the maintaining cost in process,DSCA imports Monte Carlo theory to tolerate cluster topology changes and keep the cluster unchanging at a certain probability.Simulations show that DSCA could reduce node re-association and re-clustering,the performance of DSCA is stable and experiment effect is excellent.Secondly,we design a Multi-player Game theoretic Algorithm for intra-cluster Data Aggregation(MGADA).In this algorithm,we present a novel data aggregation algorithm based on the Nash equilibrium theory by studying the inherent correlation in sensor data to analyze the competition and cooperation relationship among vehicles.What is more,it could make equilibrium analysis of cluster redundancy and cluster variation degree in VANET.We could get the optional transmission strategy by interruption process.We could reduce redundancy and aggregate accurate samples by reconstructing sub-cluster and improving aggregation environment.Simulations show that our algorithm has advantages in aggregation accuracy and topology stability.
Keywords/Search Tags:Vehicular ad hoc network, Data aggregation, Game theory, Monte Carlo
PDF Full Text Request
Related items