Estimating mutual information for the underwater acoustic channel

Estimates of upper and lower bounds on the information rates attainable for acoustic communication systems with white Gaussian sources under uncertain channel conditions are provided. These bounds account for finite duration signaling effects as well as finite coherence time and are not reliant on the presence of training data. The approach is useful in providing guidance in the selection of coding strategies based on available probe data or from propagation modeling. For the case of the coherence time of the channel exceeding that of the packet duration the estimators are computationally efficient with the use of the Levinson recursion. The reduction in the information rate relative to the known channel case is qualitatively assessed. For scenarios where the prior channel covariance is known exactly the exact bounds on the information rates are computable else the estimators provide confidence intervals for these rates. These estimators are tested on 18 kHz at sea data collected from north of Elba Italy on a 1.8 m aperture vertical array. The effects of coherence degradation as well as multi-path spread are quantified.