Efficient closed-form estimators for multistatic sonar localization

Abstract

This paper addresses the multistatic sonar localization problem using time measurements only or with bearings in the presence of Gaussian transmitter positions, receiver positions, and propagation speed variations. We develop efficient algebraic solutions to this nonlinear estimation problem through parameter transformation and multistage processing. Both situations of known and unknown statistical distribution of the propagation speed are considered. Analysis supports their performance in reaching the hybrid Cramer-Rao lower bound (CRLB) over the small error region.