Data Broadcast Scheduling In A Mobile Environment And Indexing Techniques

At present, data broadcast is an efficient way to simultaneously disseminate information to a large number of mobile clients. Under such environments, servers cyclically broadcast popular data items through the common channel; mobile clients tune in to the broadcast channel, and receive their own interested data items. Compared with the way of point to point data transmission, data item can be sent to all users who simultaneously access the data through data broadcast, and the cost of dissemination has nothing to do with the number of mobile clients, so data broadcast can meet a huge number of users to access information simultaneously, it has high scalability. There have some applications using data broadcast to disseminate information, including access stock activities and traffic conditions, and it has a broad prospect.Access time (Access Time, AT) and tuning time (Tuning Time, TT) are the two performance metrics in data broadcasting system, they are used to measure the efficiency of accessing and the energy efficiency, while the broadcast scheduling algorithms and indexing technology focus on the optimization of the two performance. This paper focuses the research on multi-item scheduling algorithms based on transaction and the index strategy of the heuristic multi-disk broadcast.Multi-item broadcast is an important way of data broadcast, while the main research is how to organize the sequence of data items in order to minimize the average access time of users. This paper introduces A algorithm of the artificial intelligence, it is implemented to optimal scheduling of multi-item broadcast, we design the cost function and the optimal searching algorithm, and then set a window size to find the optimal solution for the algorithm much more quickly. The algorithm not only has a good performance for the high accessing frequency of users, but also considered the less accessing frequency of users, eventually generates the optimal sequence of data broadcast. Experimental results show that the algorithm can effectively reducing the average access time, it will have better performance when the mode of requests is skewed.Tree index is one of the indexing technology of data broadcast, it can significantly reduce tuning time, so it is mostly researched extensively. We design an SIHMD strategy for the heuristic multi-disk broadcast. According to the difference of access probability among disks and the similar access probability of data items within a disk, we establish a balance index tree for each disk, while using VF algorithm to establish a skewed index tree for the root of each disk. Thus greatly reducing tuning time of the popular data items, at the same time, the less popular data items'tuning time is not increase obviously. In order to further optimize access time, we also design the algorithm of broadcast index tree. In a broadcast cycle, the popular items are distributed more index nodes. Experimental results show that the algorithm not only achieves the optimal tuning time, but also keep access time within a reasonable limit, achieved the better compromise between access time and tuning time.