Some speculations on source localization in uncertain ocean environments

Abstract

Incomplete environmental knowledge is a factor that can adversely affect the prediction capability of physics-based signal processors to localize acoustic sources in ocean waveguides. Environmental uncertainty effectively introduces unwanted additional degrees of freedom into the signal processing scheme. Including a quantitative measure of this incomplete information into the processor can be accomplished by considering the influence of alternative environments, each weighted by their probability of occurrence. An approach is outlined to account for environmental uncertainty in matched field processing by the use of polynomial chaos basis expansions to represent the replica vectors in terms of stochastic Green’s functions. These bases are interpreted as representing the effect of introducing additional degrees of freedom due to environmental uncertainty, because the pressure field replicas are specified as random processes. In this manner, a processor correlates the measured pressure field with a set of statistically weighted propagators between a replica source location and a sensor array. [Work sponsored by ONR through NRL base funding.]