Robust matched-field processing in uncertain shallow-water environments using an Lp-norm estimator

An optimal approach to matched-field source localization in the presence of environmental uncertainties is the maximum a posteriori (MAP) estimator. The MAP estimator can be interpreted as an exponentially-weighted average over environmental realizations. In practice, only a finite number of environmental realizations can be included in this average resulting in a suboptimal processor. We propose an L_p-norm estimator as a robust alternative to MAP in the presence of finite environmental sampling. We also show, using wavenumber gradients, that accurate localization estimates can be obtained using environmental realizations besides the precise true ones. Simulation results from a shallow-water environment are presented to illustrate the performance improvement