Skip to Main Content
The prevailing pulse compression technique utilises a matched filter receiver in order to achieve the highest attainable output signal-to-noise ratio (SNR), and relies on a judicious choice of transmitted waveform to achieve range response with narrow mainlobe and low sidelobes. As is well known, the magnitude of the frequency response of the matched filter is identical to the magnitude of the spectrum of the transmitted waveform. The logic is to amplify those frequencies where the signal's power spectral density is high and attenuate frequencies dominated by noise. A recently published paper [M. Shinriki et al., 2006] suggests a new pulse compression approach that is the antithesis of the matched filter. Rather than utilising the power in the spectral mainlobe of the transmitted waveform, the suggested mismatched filter attenuates the frequencies occupied by the waveform's spectral mainlobe and enhances the sidelobe frequencies. In this way, the suggested approach squeezes a 1 ms delay resolution out of an unmodulated 5 ms pulse.