移动社会网络中基于多维上下文匹配的数据转发算法

通过研究移动社会网络中的多种上下文信息对节点移动模式的影响,提出了基于多维上下文认知的数据转发算法MCMF。该算法综合考虑物理邻接性、社会相似性以及社会交互性3个维度的上下文信息来进行动态数据转发决策。首先消息携带者节点通过物理邻接匹配获得邻居节点集合；然后通过社会相似性匹配在邻居节点集合中选出候选节点子集,并基于社会网络的社群特征,采用马尔可夫预测方法在候选节点子集中选出最优中继节点；最后设计高效的数据转发算法。仿真实验表明,相比于其他3种著名算法,该算法在交付比率和开销比率方面具有较好的性能。

Data Forwarding Algorithm Based on Multidimensional Context Matching in Mobile Social Networks

Through studying the effect of a variety of context information on the mobility patterns in mobile social networks,this paper proposed a multidimensional context matching forwarding (MCMF) algorithm.In this novel algorithm,three dimension contexts,which are physical adjacency,social similarity and social interactivity,are used to make routing decisions dynamically.Firstly,message carrier obtains its neighbor node sets by using physical adjacency matching at the present moment.Then social similarity matching is used to search relay candidate subset of neighbor node sets,and the discrete-time semi-Markov prediction model is used to determine the best relay node.At last,the efficient data forwarding algorithm is designed.Simulation experiments based on real traces show that the proposed MCMF algorithm is more efficient in terms of maximizing the delivery ratio and minimizing the overhead ratio than other three state-of-the-art algorithms.