Cache optimizatio Hot-Point Proxy (HPProxy) using Dual Cache Replacement Policy

In recent years, the interest of Internet users turned into watching videos such as video-on-demand, online movies, news, sports, entertainments and online learning etc.,. So, the researchers are working towards the improvement of performance in media playability using proxy caching schemes. Many proxy caching schemes are developed to partially caching the video objects. Even the cache size of proxy server is very large it is limited to number of video objects to be cached at a time. Thus, the proxy cache must be optimized when more number of video objects need to be cached. This cache optimization can be achieved using good cache replacement policies. In this paper, we propose a Dual Cache Replacement Policy (DCRP) on Hot-Point Proxy (HPProxy) caching. The HPProxy works on all kind of objects and best suitable for the objects which has no popularity value. Our DCRP works based on two phases; Least Recently Loaded and Least Weighted Objects (LRLWO) and Least Weighted Sectors (LWS). The LRLWO worked as first phase on objects which are loaded least recently (oldness of the object) and Least weighted Objects. The LWS worked as second phase on sectors (a measureable unit of video object) which have low access pattern percentage and low download time. The event driven simulation results show that the dual cache replacement policy produces good cache optimization on HPProxy. Our methodology is best suitable for popularity based and non-popularity based cache optimization.