Approximate Maximum Likelihood Estimation in Split-Dwell Pulse-Doppler Radar

An approximate maximum likelihood estimator (MLE) is presented for computing velocity estimates in pulse-Doppler radar. We consider the special case of "split-dwell," or "staggered pulse repetition frequency" acquisition, in which two sequential coherent processing intervals employ different pulse repetition frequencies, resulting in two aliased velocity estimates that are then jointly unwrapped. In numerical simulations, the proposed estimator is shown to closely approximate the performance of the MLE with reduced computational complexity. The simulations also demonstrate improved estimation performance when compared to traditional velocity unwrapping via pulse pair processing coupled with a clustering algorithm.