| Summary: | 碩士 === 國立臺北科技大學 === 資訊工程系所 === 96 === With the emerging of technologies on Global Positioning Systems and wireless communications, it is possible for people to access information related to their locations any time any where. In such a wireless mobile environment, data-broadcasting provides an effective way to disseminate information to a large group of users. Two measurements are usually considered when applying data-broadcasting to provide services, the latency and the tuning time. The latency is the time between the issuing and completion of a query and stands for the quality of service. The tuning time measures the actual time when a client listens to the broadcast and presents the power consumption. In this paper, we consider to use data-broadcasting to provide the shortest path service in wireless mobile networks. We consider that the mobile clients have the memory and bandwidth restrictions. In our approach, the mobile client does not need to store the entire road network and can derive the shortest path by selectively tuning into the broadcast to save energy. On the server side, we study how to partition graphs into clusters and schedule on channel for path query. It allows the client to filter unnecessary portion when executing the shortest path query. On the client side, we provide a heuristic to compute the shortest path by listening to the broadcast. As the experimental results show, it performs well in terms of the latency and tuning time and is plausible when the client has only limited memory. In disk-based environments, we study disk page I/O performance in different graph clustering strategies while processing path queries. We also propose an alternative graph clustering strategy to compare with existing approaches.
|
|---|
