Design of mismatched filters for oversampled signals

Most of the prior work in the field of sidelobe suppression filtering for radar pulse compression focused on the design of mismatched filters for binary codes. However, in practice these binary codes are spectrally shaped to obtain analytic signals. The spectrally shaped binary code can be sampled at as low as the Nyquist rate, but in practice it is oversampled to avoid aliasing. The spectral shaping determines the necessary bandwidth and hence the sampling rate for the waveform. One of the spectrally efficient modulation techniques is based on quadriphase coding. As this spectrally shaped waveform is analytic, design of the mismatched filters in complex domain is needed. In this paper, application of minimax constrained complex LMS algorithm is proposed for the design of mismatched filters for oversampled signals.