Skip to Main Content
With the consideration of network scalability, complexity, and resource limitations, cooperative nodal decision-making in wireless sensor networks (WSN) becomes increasingly important. In this paper, a novel Martingale approach to dynamically monitoring overall network health, based upon the exploitation of a Quality of Service (QoS) requirement, is proposed. Incorporating a Martingale framework for nodal response as an integral calculation of current network statistics allows local nodal decisions to become progressively more accurate as the network life increases and the rigid constraints to power become inescapable. Hence, nodes become increasingly aware of the network conditions, based upon the current environment as well as a priori statistics, and have the ability to react with more foresight. Simulations and analysis show that this framework for sensor nodal decisions has several advantages compared with an existing learning automation approach in different fading channel environments.