NDN中一种基于数据响应时间的动态概率存储策略

A Dynamical Probability Caching Policy Based on Response Time of Data in Named Data Networking

在命名数据网络中,链路拥塞或节点负载较高会造成数据的响应时间较长,即节点由发出请求与接收数据的时间间隔较大,从而影响用户获取数据的时间,针对这一问题,提出一种基于数据响应时间的动态概率存储策略.节点根据估计响应时间与容忍响应时间,存储响应时间较长的数据,使用户下一次能够从本节点获取数据,降低数据的响应时间.仿真结果表明,本策略能够有效地提高网内的存储命中率,降低用户获取数据的时间以及缩短用户获取数据的距离.

In Named Data Networking,response time of data,which is the time interval between the time at winch request ＇s sent and data is received by node, can actually reflect status of links and nodes. Users will take a long time to fetch data from congestion links or nodes under heavy load. In order to solve this problem, a dynamical probability caching policy based on response time of data is proposed. Nodes choose to cache the data of which the response time is too long according to the estimated response time and tolerance response time. The simulation results show that the proposed caching policy can increase in-network cache hit rate, reduce the time and distance for users to fetch data.