Mobility Markov Chain and Matrix-Based Location-Aware Cache Replacement Policy in Mobile Environment

Abstract

In the location-aware services, past mobile device cache invalidation-replacement practises used are ineffective if the client travel route varies rapidly. In addition, in terms of storage expense, previous cache invalidation-replacement policies indicate high storage overhead. These limitations of past policies are inspiration for this research work. The paper describes the models to solve the aforementioned challenges using two different approaches separately for predicting the future path for the user movement. In the first approach, the most prevalent sequential pattern mining and clustering (SPMC) technique is used to pre-process the user's movement trajectory and find out the pattern that appears frequently. In the second approach, frequent patterns are forwarded into the mobility Markov chain and matrix (MMCM) algorithm leading to a reduction in the size of candidate sets and, therefore, efficiency enhancement of mining sequence patterns. Analytical results show significant caching performance improvement compared to previous caching policies.