Phase-Shift-Based Time-Delay Estimators for Proximity Acoustic Sensors

摘要

Time-delay estimation is important in a wide range of applications in oceanic engineering. In this paper, we present a novel time-delay estimation algorithm based on maximum likelihood theory for the case that the measurements are corrupted by colored or nonuniform zero-mean Gaussian noise. It turns out that the likelihood function associated with the problem is highly oscillatory, and we propose a computationally efficient technique to maximize this function. Our algorithm first obtains an initial estimate based on a smooth approximate cost function, and then refines this estimate based on the true cost function. Simulation results show that our estimator outperforms a traditional phaseshift based estimator, and that the estimation error approaches the Cramer-Rao bound (CRB) when the signal-to-noise ratio (SNR) increases without bound.