Site-role based greedydual-size replacement algorithm

Web caching is an effective Web performance enhancement technique that has been used widely. Web caching aims at reducing network traffic, server load, and user-perceived retrieval delays by replicating popular content on proxy caches that are strategically placed within the network. The replacement algorithm is the key to Web caching. This paper introduces a novel cache replacement algorithm, Site-Role Based GreedyDual-Size (termed SRB-GDS), which is a generalization of GreedyDual-Size. We argue that each website has different roles on different proxies. The proxy residing at the network edge has special characteristic that reflects the common interests of the community behind the proxy. According to its business, a community has its own interests, thus having its own favorite websites. However, this information is seldom incorporated in cache replacement algorithm policies. In addition to the consideration of the recently passed time, the size and the access cost of the document, the site role on proxy is also taken into account in SRB-GDS. Our trace-driven simulation experiment results have shown the superior performance of SRB-GDS when compared to other Web cache replacement policies proposed in the literature.