基于历史信息预测转发概率的DTN路由算法  

Design of DTN Routing Algorithm Based on Probabilistic Delay

作　　者：薛静锋[1] 范志安[1] 李建胜[2] 樊秀梅[2] 

机构地区：[1]北京理工大学软件学院,北京100081 [2]北京理工大学计算机学院,北京100081

出　　处：《北京理工大学学报》2011年第1期49-53,共5页Transactions of Beijing Institute of Technology

基　　金：国家自然科学基金资助项目(90604012);国家"八六三"计划项目(2007AA01Z220);北京理工大学基础研究基金资助项目(20071142005)

摘　　要：为了尽量减少网络先验知识的使用,提高消息转发成功率,提出了一种基于历史链接信息预测转发概率的算法,综合运用了冗余复制和相遇预测的转发策略,将消息逐步向到达目标节点传输预测概率较高的节点转发,通过二分法复制策略来增加消息转发成功的可能性,并采用了主动冗余消息删除机制.仿真实验表明,随着缓冲区的增大和节点规模的增大,该路由算法的性能明显优于Epidemic,PRoPHET和BSW算法,并且具有很好的网络适应性能.In DTN, the current routing forwarding strategies include flooding method based on copy and forwarding method based on knowledge. In order to deliver more messages with less knowledge oracle, a new routing algorithm which calculates the delivery predictability using the history contact information is proposed. The algorithm combines copy strategy and probabilistic routing. When one node meets the other node, a message is transferred to the other node if the delivery predictability to destination is higher at the other node. Binary mode copy strategy is used to distribute a message to a number of nodes to increase transmission rate and an active mechanism is used to delete redundancy messages. Simulation result shows this algorithm has a higher message delivery ratio than Epidemic, PROPHET and BSW. In addition, this superiority scales better with the buffer size and the number of nodes.

关 键 词：容迟网络 路由算法 转发策略 预测 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]