A novel adaptive learning method for low-sidelobe step frequency waveform designing

For a step frequency waveform, high range resolution can be achieved by inter-pulse sysnthesis processing and its sidelobe level can be controled by weighting in frequency domain. Howerver, high range sidelobe arise when the frequeuy steps is sparsely spaced. For this, low sidelobe can be obtained in two ways. One is by signal processing after receiving. And the other is to design the waveform with low sidelobe. In order to avoid the SNR loss, The latter is more effective. In this paper, a novel adatpive learning method is given. The method can adaptively adjust the frequencys by max sidelobe discrepancy between the matched filtering out of present and the ideal one. The result of simulation shows that it can approach the optimal frequencys for which the max sidelobe is lowest.