信息中心网络支持用户自产生业务的联合缓存路由策略

Joint Cache and Routing Policy Supporting User Generated Business in Information Centric Networks

针对用户自产生业务内容小,数量多的特点,设计了基于渐进式概率缓存和导向性副本通告的捷径路由联合策略JCRP。沿途节点以概率的方式执行缓存决策,随着数据包的下行传输,缓存概率不断减小,越靠近数据源侧的节点,存储应答内容的概率越大;并将网络节点按照兴趣请求相似度聚类为若干社区,在社区内部实现缓存副本内容的导向性通告,并建立到达缓存副本的捷径路由。仿真实验表明,JCRP算法缩短了内容请求的平均响应时延,减少了服务器响应率,有效降低了缓存副本通告的流量开销。

The characteristics of the user generated business are small content and huge quantity. In view of that, joint cache and routing policy （JCRP） based on the incremental probability caching and supervised notification for shortcut routing is designed. Nodes along the way execute cache deci- sion and cache probability falls down as the packet transmits from the content source to content con- sumer. The closer the nodes to the data source, the greater the probability of content storage. Besides, network nodes are clustered to a number of communities in accordance with the interest simi- larity, where the supervised notification is implemented within the community and a shortcut routing to the cached copy is established. Simulation results show that the JCRP algorithm shortens the aver- age response time, reduces the server response rate, and effectively reduces the traffic overhead of notification.