With the increasing events monitored in ad hoc wireless sensor networks, it generates a vital amount of data which should be effectively transmitted and stored for further retrieval and data analysis. However, since the sensor node is battery-powered and memory-constraint, it is a challenging issue for researchers on the term of reducing the energy costs and storage costs during the query processing.In this dissertation, we analyze the query processing in sensor networks. We focus on the storage-cost issue, propose a storage optimization architecture based on history cached data. We apply this architecture into the Top k query and reduce the storage costs and energy costs as well. Simulation results show that the storage dissipation can be avoided significantly while minimum energy is consumed, and the accuracy of query result can be achieved. |