Estimating Signal-to-Noise Ratio (SNR)

In many scientific studies involving evaluation of competing ocean engineering systems or the fundamental limits of system performance it is necessary to compare signal power with noise power; in particular, one needs to estimate the signal-to-noise ratio (SNR). SNR is important in the performance of detection and passive localization, where the bearing to an acoustic source is one parameter to be determined given acoustic signal waveforms received at two hydrophones. We provide a brief explanation of methods to estimate SNR. A discussion is included on the difficulty of this task both in theory and in practice. Simulation results are provided explicitly illustrating the difficulty in obtaining accurate estimates even under ideal conditions. This work has applicability to a wide variety of technical problems including the evaluation of the performance of passive localization techniques of interest to the ocean engineering community.