Synthesis of probing waveforms satisfying spectro-temporal constraints

In this thesis, several algorithms are proposed for designing radar transmit sequences satisfying temporal correlation and spectral stopband constraints. Unimodular constant modulus sequences are specifically focused on. Studies in the literature have been mostly interested in minimizing certain performance metrics such as integrated sidelobe level (ISL) and weighted integrated sidelobe level (WISL). Additionally, shaping the spectrum of the transmit waveform to avoid certain frequencies is one of the desired tasks in cognitive radars. Therefore, various algorithms have been proposed in the literature for designing sequences having low ISL or WISL values and, at the same time, satisfying some spectral constraints. In this thesis, we first utilize the genetic algorithm (GA) for designing a sequence by minimizing ISL in the frequency domain. Secondly, a new algorithm called FWISL (frequency domain WISL) is proposed to design unimodular sequences utilizing the majorization minimization (MM) method for directly minimizing the WISL metric in the frequency domain. FWISL is the first frequency domain employment of the MM method for minimizing WISL. Thirdly, we develop four more algorithms, named SMISLN (stopband MISL-new), SWPISL (stopband WPISL), SMWISL (stopband MWISL), and SFWISL (stopband FWISL), by directly minimizing ISL or WISL using the MM method in order to design unimodular sequences with suppressed power in arbitrary spectral bands and, at the same time, possessing reduced autocorrelation sidelobes. Numerical examples show that our newly proposed methods outperform some already existing methods in the literature with regard to computation time, converge in less number of iterations, and achieve better suppression in stopbands.