内容中心网络碎片化内容访问的统计机制研究

摘要

Content centric networking(CCN) significantly improves the access speed to content and reduces network traffic by introducing caching to routers.However,the caching mechanism in CCN causes the fragmented distribution of the content in the network and thus only a part of content accessing requests can be sent to the content providers (CPs) directly.It makes the CPs unaware of the accurate accessing information about their contents,thus increasing the difficulty of developing suitable pricing strategies.To solve this problem,we propose a new mechanism to ensure that CPs accurately access the count information of contents.In this mechanism,network devices are responsible for collecting the statistics of user-requested contents by adding counters of content,and sending them to the CPs once certain conditions are triggered,e.g.,threshold conditions are met or the cache is updated or expires.Experiments show that based on this mechanism,network devices can get accurate statistical information about user-requested contents and feed it back to the CPs.Therefore,CPs are capable of knowing the accurate content accessing information.%内容中心网络(content centric networking,CCN)通过在路由器中引入缓存,能够显著地提高用户访问内容的速度并减少网络中的流量.但CCN的缓存机制导致了内容以碎片化形式在网络中,只有一部分内容访问请求能够直接到达内容提供商,导致内容提供商无法真实准确地获得内容的访问统计数据,从而增加其制定恰当的内容收费策略的难度.针对这一问题,提出了一种可以保证内容提供商准确获得内容的访问信息的统计机制.在该机制中,中间网络设备中通过新增内容计数器来统计用户请求的内容信息,在满足阈值条件、缓存内容替换或过期时将内容统计信息反馈给内容提供商,提供商据此可以制定合适的内容收费策略和付费缓存策略.实验结果表明,通过本机制网络设备能够准确地统计用户的内容访问统计信息并及时反馈给内容提供商,内容提供商从而可以准确感知其内容的访问统计数据.