Abstract
Data centric storage is an effective algorithm to organize data in the sensor networks. Even though such kind of storage can save more energy than other algorithms, such as flooding based algorithm, it also wastes energy in some cases. For example, the event frequency is much more than the query frequency. To overcome this problem, we propose an index based query processing algorithm. We analyze the energy consumption of the algorithm and present conditions for the index based algorithm to save more energy than the traditional data centric storage. Finally a time-based index management algorithm is proposed to adopt the proper algorithm in different cases. Extensive experiments showed that our time-based index management algorithm can save more energy than the traditional data centric algorithm.