[IEEE Internet of Things Journal] Exploiting sparsity for underwater acoustic sensor network under time varying channels
Abstract: As a time-frequency doubly selective channel, severe multipath, Doppler as well as the large time delay characteristics of underwater acoustic (UWA) channel pose significant challenge to the research and design of UWA communication and network systems. To mitigate these negative factors, inherent sparsity contained in UWA channel has been extensively investigated to improve UWA communication via sparsity exploitation receiver. While the performance of UWA sensor network is highly dependent on that of physical layer, there is few investigations reported on exploiting channel sparsity from the viewpoint of UWA networking. In this paper, a UWA sensor network adopting sparsity exploitation physical layer is evaluated based on Network Simulator 3 (NS-3) simulation tool. The simulation time-varying channel is generated by incorporating Bellhop channel model with the statistical characteristics extracted from experimental shallow water channels. Three types of physical layer, i.e., that does not adopting sparse exploitation, adopting compressed sensing (CS) as well as dynamic compressed sensing (DCS) technique, are employed for evaluation and comparison of network behavior under different media access control (MAC) protocols. The evaluation results verify the effectiveness of sparsity exploitation in improving UWA sensor network performance at the presence of time variations, while giving quantitative comparison between enhancement achieved by the CS and DCS sparsity exploitation.