Sidelobes and grating lobes reduction of stepped-frequency chirp signal

This paper studies several methods for decreasing both sidelobes and grating lobes of FJB signal by using linear windowing, nonlinear windowing, and nonlinear stepping. We apply SVA (spatial variant apodization) as nonlinear windowing to lower down the lobes, and at the same time, keep the mainlobe unchanged. Study shows that apply SVA on single pulse of FJB burst does not help, but apply SVA on the whole FJB burst help to suppress sidelobes significantly. If we apply Hanning windowing on single pulse and SVA on the whole burst, both sidelobes and grating lobes can be suppressed significantly. At the same time, the mainlobe almost kept unchanged. At last, we propose a new nonlinear model, i.e. a tangent-function model for choosing the frequency steps in a FJB burst. This nonlinear FJB signal (NFJB) has been shown to be of much better performance than conventional linear FJB signal does. Simulation shows that if apply Hanning windowing on each single pulse in NFJB, the grating lobes can be lowered down more significantly.