A Recommended Replacement Algorithm for the Scalable Asynchronous Cache Consistency Scheme

Abstract

Acknowledging the widespread prevalence of mobile computing and the prominence of data caching in mobile networks has led to a myriad of efforts to alleviate their associated problems. These challenges are inherited from the mobile environment because of low bandwidth, limited battery power, and mobile disconnectedness. Many caching techniques were presented in literature; however, Scalable Asynchronous Cache Consistency Scheme (SACCS) that proved to be highly scalable with minimum database management overhead. SACCS was initially implemented with Least Recently Used (LRU) as a cache replacement policy. In this project we adopted the Extended LRU (E-LRU) as a cache replacement strategy to be applied in SACCS. A simulation of SACCS was done with the E-LRU, Least Recently Used, Most Recently Used, Most Frequently Used, and Least Frequently Used; and the comparative evaluation showed that SACCS with E-LRU is superior in terms of delay time, hit ratio, miss ratio and data downloaded per query.