Research On Web Hierarchy Caching Model And Cost Analysis

Web hierarchy caching model improves the cache performance, enhances the quality of service, and has been one effective way to resolve network congestion. The Web cache replacement policy can affect the performance of the Web hierarchy caching model, therefore, this paper makes a research on the cache replacement policies influencing for the cache level of Web hierarchy caching model, chooses performance metric HR and BHR, and applies cost function to evaluate the Web cache performance of different combinations of replacement policy across different levels within Web hierarchy caching model.First, this paper analyses solutions which solve the access delay of the network, focuses on the concepts and architecture of the Web caching model, and then using formal methods to describe Web caching model. This paper also summarizes and analyses existing cache replacement policy, and gives the evaluation metric for cache performance. Then, three different models of how to distribute the requests are put forward, in order to apply cost function to analyse Web hierarchy caching performance. Finally, the experiments make use of the synthetic workloads which are generated by the method of mathematic modeling, verify the authenticity of the synthetic logs, and evaluate the Web cache performance of different combinations of replacement policy (LRU, LFU, GDS) across different levels within Web hierarchy caching model.The experimental results demonstrate that compared with the other combinations, when the two-level hierarchical proxy model uses LFU-GDS combination (LFU-GDS combination indicates that lower level uses the LFU replacement policy and the upper level uses the GDS replacement policy) or LRU-GDS combination, it achieves higher caching performance in terms of HR (Hit Ratio) and BHR (Byte Hit Ratio), and receives the lower cost of the request.