Adaptive Strategy and Decision Making Model for Sensing-Based Network Applications.

Energy conservation and decision making are at the heart of wireless sensor network (WSN) research. Indeed, the limited power of sensors, which is mostly not rechargeable, accompanied to the dense deployment of sensors, which collect a huge amount of data, complicate the makers’ decision. In this paper, we propose an energy-efficient adaptive strategy and decision making technique based on a grid architecture of WSN, where a Grid-Leader (GL) is assigned for each grid. Our technique works on the three tiers of WSN: sensors, grid-leader (GL) and sink. At the first tier, each sensor applies a divide-and-conquer algorithm in order to send a reduced set of data to its appropriate GL; At the second tier, the GL combines data coming from sensors and sends useful information, i.e. satisfying a confidence threshold, about the monitored grid to the sink. The last tier, e.g. sink node, introduces two decision tables (score and early decisions) in order to make a real-time decision for each grid in the network. Extensive simulations on real sensor data demonstrated that our technique can be efficiently saving the network energy and helping in taking decisions, while maintaining an acceptable data accuracy level.