Energy Efficient Detections of Critical Events in Wireless Sensor Networks

In many mission critical applications, one of the central tasks is to obtain aggregate values, such as AVERAGE and SUM, based on which critical conditions can be detected. Immediate measures can subsequently be taken to remove associated risks timely. However, there is no successful work for detecting such critical events energy efficiently. In this paper we propose a novel approach. A fully distributed algorithm is designed, in which each sensor makes the decision purely locally. The elegance of the algorithm is that a global network behavior can be presented based on individual node actions, without a central authority to coordinate. This effectively reduces unnecessary sensor reports for network-wide coordination. Our proposed algorithm, called MEES, has the property of a zero event-missing rate, i.e., critical events are always detected. There is a trade-off between the energy cost and the false positive. Simulation experimental results show that MEES is more appropriate for large-scale WSNs with highly dynamic sensing environments. The energy cost can be saved up to 70%.