An energy efficient adaptive event detection scheme for wireless sensor network

Detection accuracy and energy efficiency are two of the most important concerns for wireless sensor networks (WSN) designed for event detection. In this paper, we consider the following fundamental problem inWSN for event detection: how to minimize WSN´s total energy consumption to provide a required detection accuracy? We focus our investigation on the design of detection scheme, which has been demonstrated recently to have a large impact on WSN´s energy efficiency. More specifically, we propose an adaptive scheme that leverages the discrepancy among individual sensor´s detection accuracy, which is obtained from a collaborated training process, to allow each sensor to operate at its most energy efficient manner while guarantees the overall detection accuracy. Comprehensive simulation results show that our adaptive scheme can improve the energy efficiency of a state-of-the-art hybrid detection method by more than 60% on large scale WSNs.