Parameter Estimation of Phase-Modulated Signals Using Bayesian Unwrapping

Parametric estimation of phase-modulated signals (PMS) in additive white Gaussian noise is considered. The prohibitive computational expense of maximum likelihood estimation for this problem has led to the development of many suboptimal estimators which are relatively inaccurate and cannot operate at low signal-to-noise ratios (SNRs). In this paper, a novel technique based on a probabilistic unwrapping of the phase of the observations is developed. The method is capable of more accurate estimation and operates effectively at much lower SNRs than existing algorithms. This is demonstrated in Monte Carlo simulations.