High-Resolution Techniques for Angle-of-Arrival Estimation

The intent of this paper, in part, is to discuss the implementation of a sampled-aperture radar facility and to address two algorithms applicable to angle-of-arrival estimation in the presence of multipath: modified forward-backward linear prediction (MFBLP) and classical maximum liklihood (ML). In the MFBLP approach, the angle-of-arrival estimation is viewed as a wave-number spectrum estimation problem. Results are presented using both simulated and real data. For the ML method, formulation of the liklihood function is developed under the provision that knowledge of the range is available and that the aperture has a multi-frequency capability. Bounds are computed, simulations of the estimation are performed, and comparisons are made with the MFBLP method. The remaining part of the paper deals with an extension of the maximum-liklihood estimation algorithm applied to non-uniform sampled-aperture arrays and the development of tight bounds on the estimation.