带有相遇概率估计的机会网络缓存替换策略

Cache Replacement Strategy with Encounter Probability Estimation in Opportunistic Network

机会网络中节点缓存资源受限,为提高节点缓存利用率,提出一种带有相遇概率估计的缓存替换(CREPE)策略.综合考虑节点的活跃度及有效相遇指数,以分布式的方式估计节点间的相遇概率,并感知给定消息在网络中的扩散程度,动态估计消息的重要程度,以确定消息的替换优先级.结果表明,所提出的缓存替换策略能够有效提高节点缓存利用率,改善网络的投递率、时延与负载性能.

In buffer a resource limited opportunistic network,a cache replacement strategy with the encounter probability estimation was designed in this paper to improve the buffer utilization.Taking into consideration the node active degree and effective encounter index,the encounter probability of nodes was estimated in a distributed manner,and the spreading degree of a given message was determined to dynamically evaluate the message important degree and to further set the replace priority of the message.Numerical results show that the proposed cache replacement strategy can effectively enhance the buffer utilization and improve the delivery probability,latency and overhead performances of opportunistic network.