SQS: An Experimental Scalable Network Structure for Efficient Querying of Micro Sensors

Wireless sensor networks operate in unattended and untethered mode where power consumption is of greatest importance. General purpose adaptive infrastructures are useful for dissemination of user interests and collection of sensor data. We present a sensor querying mechanism (SQS) which provides local query optimization and in-network message concatenation for energy efficient response delivery. The protocol is designed to operate as the lowest tier of a hierarchical architecture that enables access to tiny sensors in the target domain from a monitoring application via a service aware ad hoc backbone. The paper classifies SQS queries relative to data generation characteristics and explains the components of the querying system in this context. A comparative data traffic analysis and experimental performance results of the prototype implementation on TinyOS running Berkeley mica motes are given. The results have revealed that local optimization can reduce bit transmission up to 55% and message concatenation can further reduce transmission by 25%. It has also been shown that application specific quality of service parameters can effectively be used to construct sensor querying networks dynamically.