An Energy-Efficient Multimode Transmission Scheme for Underwater Sensor Network

The underwater sensor network plays a critical role in the ocean data collection owing to its capability of providing reliable and wide communication coverage. How to assure the lifetime performance of the network with a limited energy supply has always been a key research issue. In this article, we propose an underwater acoustic sensor network model which supports data delivery from each underwater sensor node (USN) to the sea surface sink node (SN) in either direct or relay-assisted transmission mode, i.e., transmission mode between each USN and the SN can be dynamically selected according to network's current state. To optimize its lifetime performance, we model the mode selection and resource allocation issues jointly into a nonconvex and mixed-integer programming problem. In order to efficiently solve it, we further divide the original problem into two subproblems: 1) resource allocation subproblem and 2) joint mode and relay selection subproblem. We prove that the reformulated nonconvex resource allocation problem can be equivalent to a convex optimization problem, and its optimal solution can be found by using Lagrange dual decomposition method. For the second subproblem, some critical conditions are analyzed to determine optimal relays with the criterion of balancing energy consumption between USNs, and a matching approach is designed to obtain the mode and relay selection results based on USNs' priorities. Simulation results validate that the proposed transmission scheme is effective and superior to the traditional time division multiple access scheme.