摘要

Data accessing is one of the most important issues in mobile computing. A common technique used to improve the performance of data access is caching. Due to high bandwidth of P2P channel, cooperative caching, which allows the data sharing among multiple devices, can further exploit the potential of caching technique, especially for resource-constrained mobile device. In this paper, we propose a Predictive Cooperative Cache Scheme (PCCS) which guides the cache replacement according to the value of data items calculated by means of mobility prediction. We develop a novel probability-based cost model for each data item by taking into account a mobile device s predicted movement. Base on this cost model, we implement an algorithm RangeSearch that can return the objects located within a certain range/region. Unlike the conventional optimizing scheme, PCCS reduce the cost of the communication between mobile devices and the server as well as the query delay. The simulation results show that PCCS have better performance than existing conventional cache strategy LRU.

全文