Effects of Cooperation Policy and Network Topology on Performance of In-Network Caching

In-network caching is a key component in information-centric networking. In this paper we show that there is a tradeoff between two common caching metrics, byte hit rate and footprint reduction, and show that a cooperation policy can adjust this tradeoff. We model the cooperation policy with only two parameters - search radius r and number of copies in the network K. These two parameters represent the range of cooperation and tolerance of duplicates. We show how cooperation policy impacts content distribution, and further illustrate the relation between content popularity and topological properties. Our work leads many implications on how to take advantage of topological properties in in-network caching strategy design.