Efficient Range-Doppler MMSE Estimation for Pulse Agile Radar

Radars that employ pulse-agile transmit waveforms must contend with coupling introduced between the range and Doppler dimensions, known as range sidelobe modulation (RSM). With traditional matched-filter-based range-Doppler processing, RSM appears as an increased noise-plus-interference floor that is proportional to the scattered energy in the scene, resulting in a degraded signal-to-interference-plus-noise ratio (SINR). In this work, the minimum mean squared error (MMSE) estimator is formulated for the problem of range-Doppler estimation and is shown to effectively suppress RSM. In this way, the MMSE estimator may trade mismatch loss (MML), or signal-to-noise ratio (SNR) degradation, for improved SINR in signal limited scenarios. To enable MMSE range-Doppler estimation in practical radar scenarios, a computationally efficient frequency-domain formulation is proposed, which leverages the spectral inverse filter method. The efficacy of this efficient MMSE estimator is demonstrated in simulated and open-air experimental data and is shown to improve SINR, relative to the matched filter, in signal limited scenarios.