Underwater Wireless Sensor Networks are characterized by their harsh channel conditions, lack of access to terrestrial services such as GPS for localization, limited battery life and low memory and processing capabilities. Furthermore, the effect of passive node mobility produces dynamic network topology and decreases localization accuracy. For terrestrial networks, access to GPS signals can provide accurate location information which can be used to ensure efficient network reconfiguration. Such a luxury does not exist in the underwater environment. In this paper, we propose a cluster-based network that can reconfigure as the network spatially evolves. The clustering algorithm is based on the swarming patterns of birds, fish and insects called murmurations. We reduce the murmuration behavior into two variables: orientation and opacity. The orientation data determines which nodes should be clustered together, and opacity data determines the optimal cluster head. Our results show that our murmuration inspired clustering algorithm manages network topology changes caused by underwater currents effectively and is shown to improve network lifetime by decreasing the number of cluster redefinitions and cluster head selections, thereby reducing unnecessary overhead in an already energy constrained environment. Furthermore, our results show that the decrease in network coverage due to node death is minimized by moving energy consumption to nodes with overlapping coverage regions.
|Name||IEEE International Conference on Communications|