Model-based multifrequency array signal processing for low-angletracking

Tracking low-altitude targets over the sea is problematic because of interference between the direct and reflected signal. Standard monopulse trackers can experience large errors because of multipath maximum likelihood estimation (MLE) has been used to more accurately estimate the target height in the presence of multipath MLE is a model-fitting technique where the model parameters are chosen to maximize the likelihood function. It is shown that the type of observation model has a large effect on performance. Tracking performance is compared using three different observation models employing varying amounts of a priori information. Results are presented for different array sizes: eight and 32-element arrays and two-element subarrays typical of phase monopulse. Performance is compared with that of standard techniques such as Fourier beamforming and phase monopulse