Single Channel Blind Source Separation Algorithm Based on Lifting Q-shift Dual-Tree Complex Wavelet Transform
Shexiang Ma, Liuyi Yang, Xin Meng · 2019
In order to solve the problem of frequency aliasing and time consuming of traditional wavelet transform in the single channel blind source separation algorithm, a new algorithm based on lifting Q-shift dual-tree complex wavelet transform (DTCWT) is proposed in this paper. Q-shift DTCWT can solve the problem of wavelet transform that the structure symmetry is not good and the frequency response is not identical. In addition, the lifting algorithm converts Q-shift DTCWT into time domain, simplifying the computational complexity and saving the running time. The experimental results show that the proposed algorithm can shorten the running time of the program and improve the timeliness of the single channel blind source separation algorithm.