Waveform design with stopband and correlation constraints for cognitive radar

One of the main objectives of cognitive radar is to adapt the spectrum of transmit waveforms to certain needs, such as avoiding reserved frequency bands or narrowband interferences. Besides spectral requirements, good correlation properties of the transmit waveforms are also desired in specific applications, such as range compression. Moreover, practical hardware constraints usually require the transmit waveforms be unimodular (i.e. only phase-modulated). In this paper, we propose a new algorithm named SCAN (stopband cyclic algorithm new) to design unimodular sequences with spectral power suppressed in arbitrary bands and with low correlation sidelobes as well. The SCAN algorithm, which starts from random initializations, can generate many sequences possessing similarly good properties. Furthermore, the SCAN algorithm is based on FFT (fast Fourier transform) operations and thus is computationally efficient, which facilitates long-sequence design and real-time waveform update.