A Caching Placement Strategy Based on Dynamic Router Hierarchy for Named Data Networking
Haotian Wang, Mingzhu Zhao, Meiju Yu · 2023
Named Data Networking (NDN), as one of the most promising future Internet architectures, can effectively relieve the bandwidth pressure caused by the rapid growth of network users and network traffic. It uses a content-based communication framework, in which the routers in NDN have the capacity of caching contents passing by and provide data services for incoming requests. The caching strategies, however, currently widely used in NDN are facing several challenges including low cache utilization, long service response time and so on. To address the above issue, considering the dynamic changes of routers’ location in different request/reply processes, a caching placement strategy based on dynamic router hierarchy (DRH) is proposed to determine the caching routers on the data return path. The strategy includes a dynamic hierarchical algorithm of routers, a content popularity classification algorithm and a cache-matching algorithm. By analyzing the users' requests for different contents and the location of routers, DRH determines the content popularity and importance of routers, and then caches the contents of different popularity levels on the corresponding routers. The level of each router is determined only in each request/reply process, that is, the same router has different levels in different request/reply paths. We implement DRH and conduct performance evaluations based on ndnSIM. The simulation result shows that DRH outperforms baseline solutions in terms of cache-hit rate, network throughput and average cache-hit distance.