Mobile Database Data Prefetching System Research And Implementation

Data broadcasting technology and the limitation of mobile communication bandwidth in mobile database caused large data access latency; Mobile client and fixed network frequently (initiative or passivity) disconnect allows the user transaction not to get the required data items; Power constraints of mobile devices and access remote server through the wireless network will consume a lot of energy and so these restrictive features, the traditional data caching and prefetching in mobile computing environments is widely used.Data prefetching technology is the supplement to cache technology, effective use of the user access to the temporal locality and spatial locality, through a combination of current and historical user requests, it take the initiative to forecast the data content what the user will possible access to in future, then prefetching to the client local cache. The key of data prefetching technique is how to accurately and efficiently predict the data will be used of the mobile client in future and prefetch it to the cache.According to the implementation location in the network of the prefetch Divided into server, proxy server and the client-side prefetching. This paper focuses on the data prefetching technology of the client in the mobile database for the following major studies:First, according to the structures and characteristics of the mobile computing environment, obtained that do research on theoretical and practical significance of data prefetching in mobile database technology. By analyzing the traditional model of data prefetching proposed that in client of the mobile database through use the user's history records to mining association rule and obtain the correlation user access to data, and Combined with network data resources of the update frequency, the value of evaluation function by select the prefetch data-Value_based CMIP data prefetching system. Value_based CMIP data prefetching system relies on two prefetch sets:the always prefetch collection and the failure prefetch collection. Always prefetch data in the collection will be as long as it appears in the broadcast channel prefetch to the client cache. Failure prefetch occurred in the construction of a collection of user data in failure of the requested transaction, not only the failures data to the server request, but will also request the data item associated with the failure data. While for multi-client data sent the case put forward for each client request for unified data collection and then sent to the server-side uniform.Second, Combined with the characteristics of user data in mobile computing environment, through do research on the traditional association rule mining algorithm, put forwards the association rule mining algorithms that support the mobile computing. Frequent itemset mining algorithm New_Apriori algorithm by reducing the generation of candidate itemsets and the time of scanning the transaction database, improving the traditional association rule mining algorithm efficiency and reduce power consumption in mobile clients. Incremental updating algorithm for frequent itemsets-New_FUP algorithm for the location of the mobile client changes in interest due to drift, incremental transaction database for mining association rules, which more accurately predict the user will access the data. Taking into account the mobile client system resource consumption, energy usage based on the client proposed dynamic selection support and confidence for association rule mining.Third, comparative analysis of the algorithm and data prefetching System. After a comparative analysis of theoretical and experimental method, the proposed New_Apriori association rule mining algorithm can reduce the time complexity and space complexity. At the same time, New_FUP incremental frequent itemset incremental update algorithm can transactions data set from the more frequent item sets for the choice of pre-fetch data to improve the cache hit rate. Through analyzing and comparing the simulation Value_based CMIP data prefetching system with CMIP data prefetching system, you can see the proposed Value_based CMIP data prefetching system is realizable, and reduce the failure rate of the cache, reducing the data access time and Power consumption, improve the quality of service.