An efficient multi-component signals reconstruction algorithm using masking technique based on sliding window in time-frequency plane

Due to the huge computation for eigenvalue decomposition based signal synthesis method, an efficient multi-component signals reconstruction algorithm is presented in this paper. Firstly, by analyzing the inverse transformation for Wigner-Ville distribution, a fast signal reconstruction is developed using the inherent relationship between original signal and synthesized signal. Then, the smoothed pseudo Wigner-Ville distribution is used as a time-frequency masking to suppress the cross-terms, and the sliding window method in time-frequency plane is adopted to extract signals one by one. Finally, by combining the signal synthesis algorithm and the sliding window masking method, multi-component signals reconstruction can be realized efficiently and accurately. Simulation results demonstrate the effectiveness and feasibility of the proposed algorithm.