命名数据网络中基于局部请求相似性的协作缓存路由机制

该文针对命名数据网络(Named Data Networking, NDN)应答内容的高效缓存和利用问题，依据内容请求分布的局域相似特征，提出一种协作缓存路由机制。缓存决策时，将垂直请求路径上的冗余消除和水平局域范围内的内容放置进行有效结合。垂直方向上，提出基于最大内容活跃因子的路径缓存策略，确定沿途转发对应的最大热点请求区域；水平方向上，采用一致性Hash协同缓存思想，实现应答内容的局域定向存储。路由查找时，将局域节点缓存引入到路由转发决策中，依据内容活跃等级动态执行局域缓存查找，增大内容请求就近响应概率。该机制减小了内容请求时延和缓存冗余，提高了缓存命中率，以少量额外的代价换取了内容请求开销的大幅下降，仿真结果验证了其有效性。

Collaborative Caching and Routing Scheme Based on Local Request Similarity in Named Data Networking

How to efficiently cache and take advantage of largely distributed copies poses challenges to the retrieval process of Named Data Networking (NDN). On the basis of similarity in local request, a collaborative caching and routing scheme is proposed. In the scheme, redundancy elimination in vertical requesting path and collaborative cache in horizontal local scope are effectively combined on the caching decision-making. In the vertical direction, the similar community which has the highest active value along the content delivery path is calculated based on the path caching strategy. In the horizontal direction, consistent Hash-caching is implemented to fulfill the oriented cache for the requested data in the vicinity. When a retrieve is requested, the proposed scheme dynamically performs the local lookup according to the content popularity by introduction of the local cache factor into the routing process. The simulation results show that the scheme can decrease the request latency, reduce the cache redundancy, and achieve higher cache hit ratio by comparison with existing methods.