Abstract
To support large-scale sensor network applications, in terms of both network size and query population, an infrastructure with multiple base stations is necessary. In this paper, we optimize the allocation of queries among multiple base stations, in order to minimize the total data communication cost among the sensors. We first examine the query allocation problem in a static context, where a two-phase semi-greedy allocation framework is designed to exploit the sharing among queries. Also, we investigate dynamic environments with frequent query arrivals and terminations and propose adaptive query migration algorithms. Finally, extensive experiments are conducted to compare the proposed techniques with existing works. The experimental results show the effectiveness of our proposed query allocation schemes.