A template fitting approach for cognitive unimodular sequence design

In this paper, we consider the unimodular sequence generation problem, and the sequences are required to possess both spectrum compatibility in the reverberant electromagnetic environment and low sidelobe levels of autocorrelation function (ACF). We proposed a template fitting approach based on the templates which define the spectrum compatibility requirements and sidelobe level requirements. First, we established the templates based on the actual sequence-feature requirements in the changing crowed-spectrum and clutter environment. Second, given an initial sequence, the spectrum shapes and the ACF features of generated sequences of proposed algorithm will approach those of templates by iterative process. Third, the iteration process come to end if the figure of merit shows convergence or meets specific requirements. The superiority of this approach is its calculation effectiveness. Specifically, it is capable of generating sequences with required power spectrum density (PSD) and ACF behaviors within much less computation load compared with some typical algorithms in open literature. Finally, some simulation results are provided to compare the spectrum compatibility and/or sidelobe suppression performance with that of SCAN (stopband cyclic algorithm new) and WeCAN (weighted cyclic algorithm new) algorithm, which prove the calculation and performance effectiveness of the proposed approach. HighlightsA template based technique for unimodular sequence design problem is proposed.The sequence is designed to consider sidelobe suppression and spectrum compatibility.The PSD and ACF templates are defined cognitively to meet the theme requirements.The TFA algorithm induces the PSD and ACF of sequences to converge to the templates.The TFA algorithm shows effectiveness in both performance and calculation aspects.