A LOW LATENCY PROXY PREFETCHING CACHING ALGORITHM

The Web proxy cache system was deployed to save network bandwidth, balance server load, and reduce network latency by storing copies of popular documents in the client and proxy caches for the Uniform Resource Locator (URL) requests. To solve the problem of the Web's slow end-user response time, a Web proxy caching and prefetching strategy has been developed and implemented by the auother to provide the users by the information they mostly likely want to browse in user profiles. This developed strategy uses the Reverse Aggressive technique for prefetching, which was proposed theortically. This developed strategy has been implemented with different cache sizes using a Web caching simulator. The tradional caching replacement policies such as Least-Recently-Used (LRU), Hybrid, and Size policies were already existed in this simultor. This simulator has been modified by the work in this paper such that the most recent replacement policies; Last-In-First-Out (LIFO), First-Try, Swapping and Place-Holder policies under infinite sized cache have been implemented. The performance measurements of the developed strategy have been studied using the tradional replacement policies, and the most recent replacement policies. Also, a comparative study has been done to clarify the benefits of the Reverse Aggressive caching prefetching algorithm with respect to the Fixed-Horizon caching prefetching algorithm with respect to the Reduced Latency (RL). According to the implementation results, it has been found that the average latency has been reduced at a higher degree by using the Reverse Aggressive cache prefetching strategy.