Dynamic Cache Invalidation Scheme in IR-Based Wireless Environments

Abstract

Traditional cache invalidation schemes are not suitable to be employed in wireless environments due to the affections of mobility, energy consumption, and limited bandwidth. Cache invalidation report (IR) is proposed to deal with the cache consistency problem. However, the main drawback of IR-based schemes is the long latency of data access because the mobile hosts (MHs) need to wait next IR interval for cache invalidation when the cache hit happens. In this paper, we propose a dynamic invalidation report (DIR) to reduce the latency of data access when the MHs query data. DIR contains an early cache validation mechanism by utilizing the validation messages. Therefore, the MHs can verify their cached data as soon as possible. Next, we design a predictive method to dynamically adjust IR interval to further reduce the latency called DIR-AI (DIR with adjustable interval) scheme. Finally, we evaluate the performance of the DIR and DIR-AI and compare them with the existing invalidation report schemes by using NS2 (network simulator). The experimental results show that DIR reduces averagely 54.3% and 34.3% of latency; DIR-AI reduces averagely 57.35 and 38.6% of latency compared with TS (TimeStamp) and UIR (updated IR) schemes respectively.