Efficient location management techniques for moving objects in mobile environments

Abstract

Mobile computing has become a reality as a result of the convergence of two emerging technologies: the appearance of powerful portable computers and the development of fast reliable wireless networks. In this new computing paradigm, computing entities like resources and users (e.g. humans, cars) are not required to remain in a fixed position in the network but possess the freedom of mobility and portability as well as the ability to issue queries regarding other objects of interest which could also wander around. They are therefore called moving objects. The novel characteristics and abilities of these moving objects under the new computing environment have enabled an entire new promising class of applications, LDIS (Location-Dependent Information Services), operating within a mobile environment. As continuous movement is the essential feature of moving objects, location management plays a fundamental role in supporting efficient LDIS applications. From the literature study in this thesis, it has been found that despite various efforts made and achievements gained previously, there is still much uncovered room for efficient techniques to improve location management system performance. This thesis aims at combating the limitations of previous work and proposing efficient location management techniques. The design goals these techniques seek to achieve can be summarized from the view points of both moving objects and system requirements as a whole, including query awareness, movement awareness, cost optimization and error tolerance. To achieve these goals, three new location management models are proposed and evaluated, namely, the query-aware model, basic cost-based model and extended cost-based model. The proposed models fulfill the design requirements in two ways. First, the models with their associated schemes have lower communication costs (i.e. fewer update messages from objects moving in the system are needed for position tracking), which leads to lower energy consumption. Second, from the system point of view, optimal resource utilization is achieved. On the one hand, the models would lead to a lighter work load at the server side. On the other hand, they also improve the efficiency of query processing with more precise query results generated and produce a higher service satisfaction level of the system.