Research On The Data Offloading In The Vehicular Heterogeneous Networks Based On VANET And LTE

Global mobile data is booming rapidly thanks to the increasing popularity of smart vehicular terminals.And because of mobile data booming,how to cope with the cellular network data overload has drawn more attention.Especially,delay-tolerant data offloading in the vehicular heterogeneous networks based on VANET and LTE,which offload the delay-tolerant data to VANET from LTE,not only can cope with the data overload caused by the development of smart vehicular terminal,but also fit the needs of the future development of Intelligent Transportation.However,delay-tolerant data offloading under different scenarios still has many theoretical problems which have not been solved well.One is the selection problem of nodes which will receive data initially.And this is the key of single-type data offloading under nodes-collaboration scenario.The other problem is the selection problem of different type data delivered to helper nodes initially,which is the key of multi-type data offloading under limited helper nodes scenario.Therefore,this article studies the offloading problem under these two different scenarios,and two data offloading algorithms are proposed which adapt to the limitation of these two different scenarios respectively.The main contents are as follows:(1)Vehicular Heterogeneous Mobile Network Data Offloading Using Vehicle Movement Prediction(DOVP).This algorithm is proposed to solve the single-type data offloading under nodes-collaboration scenario.By combining the direction of vehicular movement with the restrictions of roads,a probability estimation algorithm for data transfer between vehicle nodes is proposed.After that,a system offloading utility function is established,which is the expectation size of data offloaded by different selection of nodes that will receive data initially.At last,a nodes set,whose value of offloading utility function is maximum,is obtained by using greedy algorithm.(2)Data Offloading Gain Based Multiple Vehicular Heterogeneous Networks Data Offloading(OGMDO).This algorithm is proposed to solve the multi-type data offloading under limited helper nodes scenario.By combining data request matrix for different nodes and probability estimation for data transfer between nodes,multi-type data offloading utility function is established.After that,it is proved that this function is submodular function by analyzing its function characteristic.At last,data delivery selection algorithm based on data offloading gain is proposed to obtain a selection set of different data delivering to different helper nodes.(3)This article uses SUMO+MOVE andNS2 simulation software to make a simulation of DOVP and OGMDO respectively.DOVP is compared with other two algorithm which are TOMP and Random Selection algorithm.And OGMDO is compared with other two algorithm which are GA and Random Delivery algorithm.The result shows that DOVP and OGMDO have better offloading rate and average delay in their respective data offloading scenarios,which proves that these two algorithm have better offloading effect in their respective scenarios.