Research On Semantic Cache Consistency Porotocol And Replacement Policy In Mobile Environment

In mobile computing environment, narrow bandwidth, frequent network disconnections and limited client resources have become a challenge of date access. So it is much necessary for client to carry on date caching. Semantic caching is a caching scheme based on semantic locality among client issued queries. Semantic cache maintains results of previous queries and corresponding descriptions. Since semantics about the cached items are stored in semantic caching, the client is able to reason from the local cache to determine whether a query can be totally answered, how much can be answered.Existent researches in the model, consistency maintenance, replacement protocol and query processing in semantic caching have many limitations, and Semantic cache in a mobile environment by application of the architecture is to a large extent limited the application of Semantic Cache. To overcoming these limitation, this paper proposes a new framework named DSM, the detailed definition of DSM and its work mode is given,and also elaborated on the function of the design and basic realization of DSM. And based on this we defined a semantic cache consistency maintenance strategy based on the DSM framework. This strategy implemented different maintenance mechanisms according to the mobile environment network diversity, under strong connections aim at minimize the response time, under the weak link in the network in order to reduce costs as the main goal of performance optimization, and special client cache maintenance will excut after the shutdown. At the same time, the detail algorithm to maintain the data consistency is given. Formalized proving and the experiment result guarantees the correctness and the validity to this mechanism.Replacement strategy is the key technology to ensure semantic cache hit rate.Different backgrounds of the applications corresponding different replacement strategy. In this paper, a PGMC is propose aim at the Location Dependent Data(LDD),the concepts of a pre-defined grid and semantic cache structure is given, and also a detailed algorithm. Based on the typical semantic cache mode of LDD, view of the current shortage of cache replacement strategy, the data of the time factor is added to be considered, with the direction of movement of mobile terminal forecasts, joined the conventional path of user detection, and the price of the data in the conventional path is revisioned to help improve the cache hit rate, shorten the response time for saving network overhead. Results from the simulation experiments, the PGMC algorithm significantly increased and improved,cache hit rate and query response time compared to the previous algorithms with greater advantages.