Neural Network based Mobility aware Prefetch Caching and Replacement Strategies in Mobile Environment

Abstract

The Location Based Services (LBS) have usheredthe way mobile applications access and manage Mobile DatabaseSystem (MDS). Caching frequently accessed data into the mobiledatabase environment, is an effective technique to improvethe MDS performance. The cache size limitation enforces anoptimized cache replacement algorithm to find a suitable subset ofitems for eviction from the cache. In wireless environment mobileclients move freely from one location to another and their accesspattern exhibits temporal-spatial locality. To ensure efficient cacheutilization, it is important to consider the movement direction,current and future location, cache invalidation and optimizedprefetching for mobile clients when performing cache replacement.This paper proposes a Neural Network based Mobilityaware Prefetch Caching and Replacement policy (NNMPCR)in Mobile Environment to manage LBS data. The NNMPCRpolicy employs a neural network prediction system that is able tocapture some of the spatial patterns exhibited by users moving ina wireless environment. It is used to predict the future behaviorof the mobile client. A cache-miss-initiated prefetch is used toreduce future misses and valid scope invalidation technique forcache invalidation. This makes the policy adaptive to clientsmovement behavior and optimizes the performance comparedto earlier policies.