The Research Of Message Transmission Prediction Algorithm Based On Community In Opportunistic Networks

Now a large number of mobile devices with low cost, short distance communication ability(such as a Wi-Fi / Bluetooth smart phone, PDA and wireless equipment) has been widely used and carried. The theoretical works of opportunistic network begin to be put into practical application. Opportunistic network is a kind of DTN, in which there may be no fully connected path between source node and destination node, and the nodes take advantage of nodes’ mobility to transmit messages. Under the condition of limited node energy, storage space and load capacity, data transmission will face great challenge. How to improve the performance of opportunistic network, such as improving the success rate of data transmission, reducing message transmission delay and the network load, is the key problem need to be solved. According to the community property of human activities, this paper proposes a kind of community-based message transmission prediction algorithm. The main work as follows:(1) The design of a mobile network model based on community. The model simulates the community of human social activities. In the community, the nodes with higher density move slowly and meet frequently; Outside of community, the nodes meet less frequently. A node moves into which community decided by its interest, so nodes may visit some communities frequently, while they seldom even never visit the other communities.(2) Put forward a distributed community division method. According to the community characteristics of nodes, the nodes are partitioned into different communities by historical mobile information. The nodes with similar interest are divided into the same the same community. This method has low complexity and adapts to the change of community structure in the network.(3) The design of message transmission algorithm based on probabilistic prediction of community. As nodes have limited movement range, which limit the scope for messages delivery. It defines a kind of transferable transmission probability to community, which is used to describe the community connectivity.The transmission probability predicts the optimal path between communities and expand the deliverable scope of messages.(4) Improve the meeting probability prediction algorithm for message transmission in the community. Nodes have higher meeting probability,which will cause excessive forwarding message by use of the traditional prediction algorithm. So it chooses only one hop nodes to forward messages. In this way, there will be less nodes participating into forwarding messages, which reduce the redundant message copies, meanwhile it is introduced a ACK mechanism to eliminate redundant copies by itself.(5) Implement simulation experiment of this algorithm from the aspects of community size, node activity, the speed of mobile nodes, and messages life cycle. By comparing with other algorithms, simulation results show that the algorithm proposed in this paper improve the success rate of message transmission, and reduce the network load in a certain extent.