Abstract:
The effects of internal waves on low-frequency (25-Hz) narrow-band source localization performance are investigated at very long ranges (1000 km). Sound-speed perturbations induced by an internal-wave obeying the Garrett--Munk spectrum are incorporated into the normal mode solution to the wave equation using the adiabatic approximation. Sensitivity of source localization on the internal wave modal content is examined. For the entire range of signal-to-noise ratios (SNR) and degrees of randomness in the internal wave field, the optimal uncertain field processor (OUFP) provides both the signal processing algorithm as well as the limitations on source localization performance. For high SNR, the localization errors are local and the performance level of the OUFP is predicted by the hybrid Cramer--Rao lower bound (CRLB). In fairly structured internal-wave environments, significant performance gains are made optimally localizing the source. Under very random internal-wave environments, the environmental variability limits the performance of all signal processing algorithms. [Research supported by ONR (Ocean Acoustics).]