Proactive Cache Placement and Optimal Partitioning in Named Data Networking

Abstract

In-network caching plays an important role in improving network performance in Named Data Networking. The effective cache management in NDN involves cache placement, replacement and eviction policies. This paper presents low complexity, efficient cache placement strategy by using cache-span concept. The popular objects are stored within cache-span radius of Autonomous system to increase the cache hit ratio. The proactive cache placement identifies router locations to store the popular contents to serve large proportion of requests within cache-span. An optimal partitioning based on cache-span level is also presented. The partitioning is required to store content objects and information of cached content of nearby routers defined in cache-span. Our simulations show that proactive caching increases cache hit ratio and reduces round trip latency as compared to existing approaches. The methods are evaluated using analysis model and also by simulation using ndnSIM. The simulation study shows that our method outperforms default NDN caching strategies and also more sophisticated approaches in terms of cache hit ratio and round-trip latency.