带有消息投递概率估计的机会网络自适应缓存管理策略

高效的缓存管理策略能够有效提高机会网络中节点的缓存资源利用率。消息的投递概率直接决定了消息的转发与存储必要性，该文提出一种带有消息投递概率估计的自适应缓存管理策略，通过构建节点连接状态分析模型，以分布式的方式感知节点服务能力，从而估计消息的投递概率，进而确定消息的转发与删除优先级，以执行缓存管理相关操作。结果表明，所提出的缓存管理策略可降低网络负载57%，并有效提高消息的成功投递率，降低消息的平均投递时延。

Adaptive Buffer Management Strategy with Message Delivery Probability Estimating Method in Opportunistic Networks

The buffer resource utilization in opportunistic networks can be improved by efficient buffer management strategy. The delivery probability of message is directly related to its necessity of forwarding and buffering. An adaptive buffer management strategy with message delivery probability estimating method is proposed. Through establishing the node connection status analysis model, the diversity of node service ability could be evaluated. Accordingly, to estimate the delivery probability of the message. Furthermore, the transmitting and removing priority could be determined reasonably to perform buffer management operations. Numerical results show that overhead ratio can be reduced about 57% by the proposed buffer management strategy. The delivery ratio is improved and the delivery latency is reduced efficiently.