A Routing Algorithm Based On Regional Friendship For Opportunistic Networks

In recent years,mobile communication has gradually become an important way for people to share information and get services.However,traditional wireless communication has to use some infrastructure(such as base statio n)to guarantee the connectivity between the communication nodes to complete communication,which make it difficult to communicate without infrastructure to deploy in the harsh environment.In order to solve this problem,a new type of network came into being –Mobile Opportunistic Networks(MON).Node location and network size change frequently and there is no need to determine whether there is a path,but making use of "encounter opportunity" which is created by node mobility to complete the communication when communicate to end in MON.So it can solve the communication problem in discontinuous network and has broad application prospects.However,the problem such as the complexity of topological structure and lack of routing information also bring tremendous challenges to message forwarding,while opportunistic routing algorithm can make the transmission and dissemination of message maximum by predicting the node moving rules.Generally,the analysis of the node’s historical trajectory information can predict the node moving rule to a certain extent,and the MON which is carried by the human can make use of the social attributes of the node mobility to do it in order to save resources and improve the accuracy of prediction.This thesis focuses on how to improve the performance of opportunistic routing algorithm by using the social attributes of node mobility.The social attributes usually include the community,the node closeness,the node status and so on in the design of social opportunistic routing algorithms.Note that the node closeness degree evaluation has impact on the accuracy of the contacting probability prediction,and thus affecting the performance of opportunistic routing algorithm using the closeness degree to predicting the node contact probability.Therefore,an efficient routing algorithm need a comprehensive analysis of the various factors on node closeness degree,and then a framework model should be established by integrating the multi-factors to design the forwarding strategy.Moreover,previous studies have not considered the influence of node location on the evaluation of node closeness.In general,the nodes in different regions maintain different close neighbors.In this thesis,the node closeness relating with both the time and the located regions is considered,we construct the closeness evaluation model exploiting the node contacting histories.Thus,the regional friends of nodes can be obtained based closeness degree in different region.Finally,an opportunistic routing algorithm based on regional friendship is presented.The experimental results indicate that the proposed algorithm can get a higher delivery ratio and a lower forwarding energy consumption than others under different deployment density of nodes and delay constraints.At the end of this thesis,the Android prototype system which can be used for point-to-point communication is implemented based on the Wi-Fi P2 P technology.The system has functions such as the neighboring nodes searching,connection establishment of neighboring nodes and data transmissions.As last,the prototype system is tested in reality,and relevant performance data is collected to support for the future improvement.