Asymptotic Results For Multidimensional Sensor Array Processing/sup 1/

The problem of estimating signal parameters from the output of a sensor array is addressed. The Maximum Likelihood estimation procedure is a systematic approach to many parameter estimation problems. The deterministic ML method has been formulated and several methods for maximizing the cost function have been proposed, However, the asymptotic distribution of the estimation error has not been reported for the general case. The distribution will be derived in this pa- per. Multidimensional signal subspace methods have been proposed recently with advantages over conventional one- dimensional subspace methods. We examine the asymp- totic properties of a weigh" subspace fitting problem. A weighting matrix is proposed which gives the subspace fit- ting method the same asymptotic distribution as the ML method.