Analytical model of hierarchical cache optimization in the network with unbalanced traffic

Abstract

Caching is important tool for reducing Internet access latency, network traffic and cost in content delivery networks (CDN). In case of hierarchical networks with several levels of caching (e.g., IPTV network), the problem is where and how much cache memory should be allocated in order to achieve maximum cost effectiveness. In this paper, we continue study of hierarchical cache optimization and present a model of hierarchical cache optimization for CDN with two levels of hierarchy and unbalanced traffic. This model depends on several basic parameters: traffic throughput for each low level node, cache effectiveness as a function of memory size, and cost parameters. Some reasonable assumptions about network cost structure and cache effectiveness function allow us to obtain an analytically optimal solution of the problem. We then analyze the factors that impact this solution.