Optimal frequency selection for energy efficient underwater acoustic networks

The underwater acoustic channel is characterized by a path loss that is dependent on both the distance and the frequency of communication. Given this dependence, it has been previously demonstrated that for a given communication distance, there is an optimal operating frequency, where conditions for signal propagation and noise are most favorable. In this work, we consider extending this optimal frequency concept to scenarios in which the frequencies that can be employed by the system are constrained. Such constraints are important considerations for practical system design. The first problem we study is to find a single frequency that minimizes the energy over a number of links of varying lengths. An approximate model for this frequency is proposed that is very close to the true optimal. We then generalize this problem to finding the best frequency band, within which the frequency can be tuned for different link lengths. We demonstrate how our model is applied to a 2-D network scenario. We simulate random node placement for such a network, and we observe that the optimal frequencies are very close to the proposed model.