A Cramer-Rao bound for source range estimation in a random ocean waveguide
Matched-field source localization methods which exploit complex acoustic multipath propagation models are particularly sensitive to errors in the assumed environmental parameters. In this paper, the Cramer-Rao Lower Bound on source range estimation performance is used to determine the importance of having the prior probability distribution of the uncertain environmental parameters. Theoretical results obtained using an adiabatic normal mode propagation model indicate that without the prior distribution of the environmental parameters, the CRLB may diverge even when the number of unknown sound-speed parameters is less than the number of acoustic modes. When the prior distribution of the environmental parameters is available, however, the CRLB indicates a significant improvement in range estimation performance is possible.
