Energy-Efficient Traffic-Aware Routing in Underwater

Abstract

Underwater acoustic networks are a special type of wireless sensor networks deployed in a harsh oceanic environment for mission critical tasks. In this unique sensor network, energy efficiency is the most critical problem. When Maximum Residual Energy Routing is adopted in actual battery-powered underwater acoustic sensor networks, further improving the energy consumption in this protocol and prolonging the system lifetime becomes a significant problem. In this study, we examine the Maximum Residual Energy Routing Protocol (MREP) and propose a new model for energy utilization by considering the relationship between successful packet sending probability and node-distance. Based on this model, we develop a new method for improving MREP. Compared with previous model, the network energy usage is more uniform in the new model and the result is an increase of 20%~30% in system lifetime. The limited bandwidth and power resources as well as the 3-D topology in underwater acoustic sensor networks have made the geographic routing a favorite choice. While most of the detouring strategies in the existing geographic routing do not work well for underwater sensor networks, the spanning tree routing detouring strategy can efficiently find a detour for a packet when greedy forwarding fails. However, the effectiveness of the spanning tree routing depends largely on the quality of the preconstructed spanning tree. Most of the existing spanning tree construction algorithms build trees in a top-down and centralized fashion and do not consider the traffic load and residual energy level in the network, and therefore is likely to create trees with poor routing performance. In this research, we propose novel spanning trees, namely Distributed Traffic-Aware Routing Tree (TART) and Distributed Energy-Aware Routing Tree (EART) which are constructed completely in a bottom-up fashion with the traffic load and residual energy level in mind. Simulation results show that those spanning trees have very few conflicting hulls, result in much higher path throughput and residual energy level when compared against other spanning trees, leading to a better routing performance in a 3-D underwater sensor network.