Study of Threshold Setting for Rapid Detection of Multicomponent LFM Signals Based on the Fourth-Order Origin Moment of Fractional Spectrum

By using Fractional Fourier Transform (FRFT) for detecting multicomponent linear frequency-modulated (LFM) signals, weak signal components are usually shadowed by the sidelobes of strong ones and cannot be detected effectively. The CLEAN technique is widely used to improve such a situation via subtracting the detected strong signal component from the multicomponent signals iteratively and applying detection algorithm to the remainder signals. Thus the adverse effect of strong components can be eliminated and weak signal components may be detected. However, since only the strongest LFM component is detected and subtracted in each iteration, the number of detection iteration will be directly proportional to the number of LFM components. Using concept of shading coefficient, a novel threshold setting method for rapid detection of multicomponent LFM signals has been proposed by analyzing the shading relation among the fourth-order origin moment of fractional spectrum (OMFrS) of different LFM components. By employing the proposed method, several LFM components whose fourth-order OMFrS at their own optimal rotation angles are not shadowed by strong components can be detected at the same time in one iteration. Consequently, the number of detection iteration and the computational complexity are reduced dramatically. Simulation results are given to verify that this new approach is valid.