The Research On Data Broadcast Scheduling Algorithm In Mobile Computing Environments

The development of mobile computing will enable users to access data anywhere and anytime. However, compared with fixed network, the wireless communication network which supports mobile computing has low bandwidth and poor quality. In order to support a large number of mobile clients accessing data in server concurrently, data broadcast is provided. It is a new data dissemination mode, where server broadcasts data in air, and mobile clients retrieve data from air. Data broadcast is an important technique to improve the scalability of mobile computing systems. It enjoys a promising future in the fields of public news dissemination and military application, etc. Data broadcast research mainly focuses on scheduling algorithm, index technologies, caching and prefetching, concurrency control. Among these topics, research of data broadcast scheduling is a key technology.This paper emphasizes on broadcast content selection strategy and broadcast scheduling algorithm for efficient access of multiple data items. We first introduce the research of data broadcast scheduling algorithms for accessing multiple data items. Through analyzing the exist algorithms, we get the needful features of multiple data items broadcast scheduling. The key elements of data broadcast model construction are analyzed, and then a real-time broadcast model based on demand driven framework is presented. An important issue in data broadcast scheduling is how to select the content of broadcast channel based on the dynamical queries from the clients. After considering the requirements of both accessing multiple data items and meeting the timing limitation of the transaction, we propose priority computing model. A new content selection algorithm based on priority named PBQS is presented, which reduces selection overload by using pruning technique.QEM method is a classical data broadcast algorithm for accessing multiple data items. Its main disadvantages are lack of consideration to data deadline, and the data item selection for broadcast is only based on their access probabilities. Therefore, that results in not working well in time-critical data dissemination. In this paper we introduce an improved algorithm named PBQEM, which applies PBQS strategy. PBQEM is an expansion method from general transaction to real-time transaction. It is validated and analyzed. The experiment results show that PBQS selects the broadcast content dynamically, which based on the current user query queue state. It has high respond ratio and low selection overload. PBQEM has good performance...