基于社区的机会网络路由策略分析-analysis of opportunistic network routing strategy based on community.docxVIP

优秀毕业论文 精品参考文献资料 Abstract As a special type of wireless communication network, Opportunistic Networks consist of mobile communication terminals carried by people, animals or transporting equipments. In Opportunistic Networks, the end-to-end paths between nodes are instability due to the node mobility, limited energy and caching, etc, the data transmissions rely mainly on the node mobility and opportunistic contacts. The classical routing algorithms in Opportunistic Networks are mostly based on the study of network nodes random mobility model, and rarely consider the community features of actual mobile communication equipment carriers, which lead to the result that the classical protocol can not run effectively. In view of the lack of transmission delay and node buffer overflow when the carrier is waiting for the intermediate nodes in PROPHET algorithm, this paper improves the PROPHET algorithm from two aspects: the controlling of message copies and the management of node cache, then a Modified PROPHET algorithm (MPROPHET) is proposed. The algorithm determines the maximum number of message copies dynamically based on the number of online communities as well as the node encounter probability, and then realizes the multi-path parallel transmission according to the utility function; at last it manages the node cache by the ACK buffer management mechanism. Simulation results show that the MPROPHET algorithm can improve the message delivery probability, reduce network overhead and message transmission delay effectively. Then a Community-Based Centrality Predicable Spray and Forward Routing Strategy (CPSFR) is proposed according to the centrality and the node transition probabilities in different geographic communities. The algorithm generates a certain number of message copies dynamically using the transition probability between communities and community number, then chooses the intermediate nodes depend on the centrality and transition probabilities until the TTL is inv