A Modified Least-Squares Mismatched Filter for Use in Radar Applications with Additive Noise

Waveforms used for passive synthetic aperture radar applications typically have poor range resolution which leads to difficulty distinguishing between small targets within a scene. This problem is magnified when there are high sidelobes in the filter response. This paper formulates a modified least-squares mismatched filter used on noisy, oversampled, OFDM signals in lieu of a traditional matched filter. This paper considers the role of signal oversampling on the filter's performance and evaluates the mismatched filter against the matched filter in various noise environments. Range profiles are presented and then image formation is carried out over a narrow aperture to demonstrate the superior sidelobe suppression of the modified least-squares mismatched filter in the presence of additive noise.