Least-Squares Optimal Mismatched Doppler Processing

Abstract

Doppler processing is a well-known tool for the discrimination of scatterers based on their rates of radial motion. It is part of nearly every kind of radar mode via the coherent combination of echoes from a set of pulses. While windowing is likewise a well-known approach to reduce Doppler sidelobes (which also incurs degraded resolution and SNR loss), here we leverage recent work on least-squares (LS) optimal mismatched filtering that employs over-sampling for high-fidelity along with the means to compensate for the attendant super-resolution degradation that would otherwise arise. The result is a mismatched Doppler processing (MMDP) transformation that reduces sidelobes with less SNR loss and resolution degradation than windowing, as demonstrated using simulation and free-space measurements.