Distributed smart sensor scheduling for underwater target tracking

Underwater Sensor Networks include multiple sensor nodes that possess the ability to sense and communicate the environmental information where they are deployed. It is desired that these networks are intelligent in the sense that they allow for rapid deployment, self-organization, energy conservation, and fault tolerance via implementation of rapid multi-objective optimization algorithms. This paper proposes a decentralized sensor scheduling approach that enables dynamic space-time clustering for an energy-efficient target-tracking sensor network. Each sensor node is modeled as a Probabilistic Finite State Automata (PFSA) that governs its energy consumption, sensing, and communication activities. This PFSA allows the sensor nodes to dynamically change their states to conserve energy when a target is absent and turn on their high power sensing devices when the target is present. The algorithm proposed is compared with traditional scheduling schemes and the results show that the proposed method conserves energy while maintaining an accurate track estimation in a decentralized manor.