Resource Scheduling Strategies For Underwater Wireless Sensor Networks

The Underwater Wireless Sensor Networks (UWSNs) have brilliant prospect in field of un-derwater pollution monitoring, ocean data collection, disaster prevention, assisted navigation and tactical surveillance. The underwater sensor nodes are powered by battery and energy harvesting is almost impossible. Therefore, energy saving is an important topic for UWSNs. Based on the National Natural Science Foundation of China Project, the paper focuses on the energy-efficient resource scheduling strategies of UWSNs.At first, for the packet communication problem in UWSNs, the intermediate nodes selection problem and power scheduling problem are investigated. The intermediate nodes selection strategy is proposed for the packet delivery between anchored nodes. For the direct communication between two underwater nodes, we proposed the Bit Error Rate based power scheduling scheme. For the communication between the base station and the underwater nodes, the power scheduling for the routing problem is investigated. The flooding area is determined through the power control. This method has short delay and low energy consumption.Secondly, considering the limited spectrum of underwater acoustic channel, the access schedul-ing problem of cognitive acoustic system is investigated. The state of primary user’s channel can be modeled as a Hidden Markov Model. The Hidden Markov Model can be trained through underwa-ter nodes’observations. The channel with highest idle probability will be sensed by an underwater node. Considering the possible adversary jammers, the anti-jamming channel sensing strategy is proposed.Thirdly, we focus on the scheduling problem of UWSNs under best linear unbiased estimator. Through scheduling the quantized bits of different sensors, the energy consumption can be reduced. When the noise variance is known, the scheduling problem can be described as a Kuhn-Tucker problem. The optimal scheduling can be reached through solving the Kuhn-Tucker problem. When the noise variance is unknown, the scheduling is reached through feedback adjustment. In the soft real time parameter estimation scenarios, we try to guarantee the soft timing constraint through adding redundant active nodes. The approach shows better performance when the delay of each hop varies largely.Finally, the node state scheduling problem for underwater multi-target localization is inves-tigated. To deal with the problem, we try to determine an area that contains the target. Then the nodes near the area will be turned to proactive localization state. Due to the space sparse property of underwater targets, the magnitude of original source in different areas can be obtained through compressed sensing. Thus, the nodes near the area with high signal magnitude can be selected to participate the proactive localization. Such node state scheduling approach is quite energy-efficient.