LAUSR

Underwater acoustic sensor networks (UASNs) provide new opportunities for exploring oceans and consequently improving our understanding of the underwater world. UASNs usually rely on hardware infrastructure with poor flexibility and versatility. Compared with wireless sensor networks, UASNs are quite expensive to manufacture and deploy. Due to the unique data format, protocols, and service constraints of various applications, UASNs are typically deployed in a redundant manner, which not only leads to waste but also causes serious interference due to the presence of multiple signals in the same underwater region. Software-defined networking (SDN) provides an innovative means of improving the flexibility of underwater systems. In this paper, we present an SDN-based UASN framework, followed by the design of a clustering method in which learning automata and degree-constrained connected dominating sets are employed. We then propose a load balancing mechanism involving multiple controllers, based on the consistent hashing algorithm. Finally, we describe a simulation program (called UASNs hypervisor) that we developed and implemented to assess the network survival time, bit-error rate, and computational complexity. The experimental results show that the UASN was improved significantly. This work provides important theoretical and technical support for the implementation of SDN-based UASNs.