Design of unimodular sequence train with low central and recurrent autocorrelations

Abstract

This study considers the unimodular sequence train design with low central and recurrent autocorrelation sidelobes. The authors attain the design goal by minimising the integrated energy in central and recurrent autocorrelations with user-supplied weights. They devise a way to employ a fast Fourier transform in calculating both the objective function and its gradient, such that the considered design problem can be efficiently solved using the general gradient method. Numeric results indicate that the proposed method can effectively suppress the range sidelobes concerned and it can design sequence train of very large sequence length and set cardinality.