A Querying Processing Architecture for Sensor Networks

Abstract — Although advances in sensor node hardware, which comprises sensors, embedded processors, and communication components, have made the large-scale deployment of sensor networks a reality, sensor networks are quite limited in resources. A sensor network includes a numerous battery-operated wireless sensor nodes, which have limited processing and storage capa-bilities, and a few base stations, which are powered PCs that are possibly connected to the Internet. Furthermore, typical sensor network applications, ranging from monitoring to military and health care, generate various complex continuous queries. The querying of sensor networks requires a rich set of abstractions, techniques, and heuristics. In this paper, challenges and design goals for query processing strategies are identified. The current strategies are compared based on the query language, the in-network processing mech-anisms, and the query optimization techniques. Moreover, the integration of local and wide area sensor networks is addressed. In particular, a query processing architecture is proposed to meet the emerging needs of sensor network applications. Additionally, future research directions for query processing are outlined. I