Research On Mobile Model And Routing Algorithm In Mobile Social Network

With the rapid development of mobile intelligent devices (e.g., smart phone, PDA, etc.), through mobility of devices establishing temporary network, which makes the communication in some areas without infrastructure becomes reality. Due to most of mobile devices are carried by human, which not only have human’s mobility, but also have a particular social relationship, leading to researchers focus on the technique of mobile social networks. This new network has a great potential application in many areas, for example, Vehicular ad-hoc Networks, Remote Area Information Transmission, Location Based Services and so on.However, in mobile social network, it faces great challenge in forwarding message due to nodes’mobility and the intermittent connection of network. However, the regularity of different devices’movement mirror these of their owners, finding a suitable route between two disconnected devices implies uncovering habits in human movements and patterns in their connectivity (i.e., frequencies of meetings, duration of a contact, etc.) to forward message. Therefore, it is a major issue to design a reasonable mobile model and routing algorithm to select a next-hop node for forwarding message.In this paper, we firstly introduce research background, significance and the concept of mobile social network, and then summarize and analyze existing mobile models and routing algorithms. Due to the characteristics of MSNs, which not only have the mobility of opportunistic network, but also have complex sociality. This paper proposes VSMM mobile model based on working day model, which accords with people’s movement in the society. It can reflect people’s mobility and the social relationship between human, it also can predict the node’s state and its position more accurately. Besides, in this thesis, we analyze nodes’social property and combine with VSMM mobile model. We found that people’s behavior has "small world" property, that is to say, in some zone there is more close relationship between nodes, so we propose a routing algorithm based on community divided. In this algorithm, we use encounter frequency and the length of meeting time to balance relationship of nodes, choose top L friends as relay nodes, and then construct relationship list to forward message. In order to evaluate the performance of our proposed CBDR algorithm, we compare it with the existing algorithms SimBet and Prophet. The simulation results show that our algorithm can achieve more effective performance in the aspect of Delivery Success Ratio, number of hops, average latency and so on.