Abstract
Mobile computing is a new emerging computing paradigm which poses many challenging data management problems. For wireless computing to be truly relevant to the users, mechanisms to disseminate answer tuples in response to relational queries must be supported. To process database queries, it may require manipulating a large portion of the base relation and it means that the wireless channel can become congested very quickly. Based on two query processing models, broadcast-based query processing models and on-demand location dependent query processing, we propose two different approaches to solve this problem. For the broadcast-based query processing model, we consider how to allocate the data objects on the broadcast channel to reduce the access latency for processing the database queries. In this approach, offline query processing and online query processing are considered. In offline processing, the query is processed after the values for all the relevant attributes are downloaded. In this case, the tuning time is fixed for a query. The access time is dependent on the order of attributes on the broadcast channel. To reduce the average access time, an efficient strategy for answering database queries in mobile environments is proposed. Experiments are performed to show that our data scheduling methods can achieve a good performance. In Online processing, the query is processed during the access of the values for each relevant attribute. The tuning time is minimal, if the access order of attributes is followed as the process order of query optimization. In this problem, we propose an efficient algorithm to determine the allocation of the data on the broadcast channel such that frequently co-accessed data are allocated close to each other. Our mechanism is based on the well-known problems named maximum branching and optimal linear ordering. Experiments are performed to justify the benefit of our approach.For the on-demand location dependent query processing model, we derive the condition for data replication and propose an associated cache replacement policy. The main goal of data replication is to reduce the network traffic generated by massive queries and to speed up the query processing. Moreover, we propose eight query processing strategies, simulate their performance according to different query types and other factors of the wireless environment, discuss the simulation results, and conclude with efficient strategies according to different conditions.The personal communications services (PCS) systems provide ubiquitous and customized services. The key issue, which affects the performance of the whole system, is the location management. Current systems group cells into location regions to reduce the location management cost which includes the location tracking cost and the registration cost. By considering users' movement behaviors, four methods are proposed to derive location regions. These methods attempt to simultaneously reduce the location tracking cost and the registration cost. Several design issues are studied by considering the user movement patterns and the call-to-mobility ratio for the users. Simulations are performed to compare the performance of these methods with the existing strategies to show the superiority of our approaches.