Multichannel Time-Delay and Signal Model Estimation with Missing Observations

In this paper, we propose a maximum likelihood (ML) estimator in the frequency domain for estimating multichannel time delay and parameters with missing observations. The missing observations are described by a random Bernoulli pattern. In this context, the ML estimator for missing observations is highly sensitive to the initial conditions and complexity of a given problem. In conventional calculations, the complexity of problems will often make it difficult to obtain the optimal results. Thus, we adopted an iterative method using a genetic algorithm (GA) to develop an ML estimator for a model signal, the time delay, and the missing probability in the frequency domain. Several simulation examples were analyzed to evaluate the performance of the proposed method. The simulation results show that the performance is significantly improved if the effect of missing observations on the ML estimator is considered.