Development of intelligent web proxy cache replacement algorithms based on adaptive weight ranking policy via dynamic aging

Abstract

Nowadays, the World Wide Web plays an essential role in our lives. It has become a great useful tool for people in all facets of life. The vast usage of the World Wide Web leads to an increase in network traffic and create a bottleneck over the internet performance. For most people, the accessing speed or the response time is the most critical factor when using the internet. Web proxy cache technique reduces response time by storing copies of pages between client and server sides. If requested pages are cached in the proxy, there is no need to access the server. Due to the limited size and excessive cost of cache as compared to other storages, cache replacement algorithm is used to determine evict page when the cache is full. On the other hand, the conventional algorithms for replacement such as Least Recently Use (LRU), First in First Out (FIFO), Least Frequently Use (LFU), Randomised Policy and etc. may discard important pages just before its use. Furthermore, using conventional algorithm cannot be well optimized since it requires some decision to intelligently evict a page before replacement. Hence, this research proposes integrated Adaptive Weight Ranking Policy (AWRP) with intelligent classifiers based on Naïve Bayes (NB), J48 decision via dynamic aging factor to form intelligent replacement algorithms called NB-AWRP-DA and J48-AWRP-DA that improves the performance of AWRP based on hit rate (HR) and byte hit rate (BHR) over LRU, LFU and FIFO algorithms. In order to enhance classifier’s power of prediction by rising classifier accuracy before integrating them with AWRP, this research proposes using automated wrapper feature selection methods to choose the best subset of features that are relevant and influence classifiers prediction accuracy. The results present that by using wrapper feature selection methods namely: Best First (BFS), Incremental Wrapper subset selection (IWSS) embedded NB and particle swarm optimization (PSO) reduce number of features and have a good impact on reducing computation time. However, Using PSO enhances NB classifier accuracy by 1.1%, 0.43% and 0.22% over training NB with all features, using BFS and using IWSS embedded NB respectively. PSO rises J48 accuracy by 0.03%, 1.91% and 0.04% over using J48 classifier with all features, using IWSS-embedded NB and using BFS respectively. While using IWSS embedded NB fastest NB and J48 classifiers are much more than BFS and PSO. However, it reduces computation time of NB by 0.1383 seconds and reduce computation time of J48 by 2.998 seconds. Moreover, experimental result of intelligent replacement algorithms shows that NB-AWRP-DA enhances the performance of web proxy cache a cross multi proxy datasets by 4.008%,4.087% and 14.022% over LRU, LFU and FIFO in terms of HR. Moreover J48-AWRP-DA increases HR by 0.483%, 0.563% and 10.497% over LRU, LFU and FIFO respectively. Meanwhile, BHR rises by 0.991%,1.008% and 11.584% over LRU, LFU and FIFO respectively using NB-AWRP-DA. And by 0.0204%, 0.0379% and 10.614% for LRU, LFU, FIFO respectively using J48-AWRP-DA.