Effective Heuristics for Flexible Spectrum Access in Underwater Acoustic Networks

In this paper, we consider underwater networks with multiple sensors reporting data to a single sink node, and advocate the use of FDMA as a channel access method. In order to assign channels to users in an efficient way, we compare a number of allocation methods that explicitly account for the underlying channel attenuation and noise model. We begin by inspecting a simpler case, where the number of channels, Nch, is equal to the number of users, Nu, and shortly describe a known algorithm which finds an optimal channel allocation in a max-min fair sense. We then assume that Nch > Nu, and explore how the channel assignment problem changes. We finally propose suitable heuristic methods that allow to find effective solutions without the complexity of an exhaustive exploration of all possible assignments. Performance evaluation by means of simulation shows that our heuristic methods maintain fairness among users, while achieving a better minimum performance than optimal methods for the Nch = Nu case. We also show when it is more effective to use a given allocation method as a function of the network area extension. Finally, we compare our FDMA-based allocation method with the ALOHA random access control mechanism, showing how our channel allocation can improve the overall throughput of the network.