Research On VANET Data Collection In Urban Scenario

In recent years,with the development of society,urban traffic congestion,frequent traffic accidents and other issues have followed.As a special network of Mobile Ad-hoc Network(MANET),Vehicular Ad-hoc Network(VANET)can use the original sensory data of smart vehicles to share various types of traffic information.And this is the basis and premise of many applications in Intelligent Transportation Systems(ITS).However,VANET data collection faces great challenges due to the relatively higher speed and the frequently changed topology.Therefore,the research on VANET data collection in urban scenario is of great significance for solving the urban traffic problems.(1)When using the storage-forward method among smart vehicles for data collection,the high-speed movement of vehicles causes the intermittently connected of communication link,then makes a relatively lower data collection rate.Aiming at this problem,two different algorithms,which can establish the dynamic collection routing tree in the limited communication scene and unlimited communication scene are proposed.The algorithms utilize the real-time traffic information to dynamically determine the data transmission strategy in the data collection process,and recursively search for the parent nodes based on the idea of greedy algorithm to construct a dynamic aggregation routing tree.At last,comparative experiments on simulated trajectory datasets show that the algorithm can improve the data collection rate.(2)When using the storage-forward method among smart vehicles for data collection,part of data is carried by vehicles during the transmission,resulting in a longer data transmission time.Aiming at this problem,a road side unit(RSU)assisted data collection strategy is proposed.We firstly use the improved binary differential evolution algorithm to determine the optimal deployment locations of RSUs,and then make the deployed RSUs participating in the data collection.In the RSU deployment algorithm,the initial population is generated based on the opposite-based learning.And the individual reparation and optimization algorithm is introduced to repair and promote the potential solutions.Besides,the solution optimization algorithm is executed to improve quality of the optimal solution.After the comparative experiments in synthetic map and realistic map,results show that the collection strategy can ensure the datacollection rate and reduce the data transmission time.