Novel Nonlinear Signals Separation of Optimized Entropy Based on Adaptive Natural Gradient Learning
Without knowing the signal probability distribution and channel, novel blind source separation (BSS) of singular value decomposition (SVD) with adaptive minimizing mutual information is proposed to extract mixed signals. Adaptive natural gradient decent algorithm attains fast convergence speed and reliability. We focus on applying cost function BSS and SVD to achieve the solution of decomposition signals. The results indicate that the SVD combining minimizing mutual information can predict the extent of mixed signal and searching direction. The simulation illustrates that the method improves the performance, convergence and reliability. The different results can be attained by distinctive nonlinear function. The algorithm of adaptive changing de-mixed function is a better way to break through the limitation of nonlinear BSS.
KeywordsMutual Information Singular Value Decomposition Independent Component Analysis Independent Component Analysis Blind Source Separation
Unable to display preview. Download preview PDF.
- 1.Zhang, L., Amari, S.: Estimating function approach to multi-channel blind deconvolution. In: Circuits and Systems IEEE APCCAS 2000, vol. 12(4-6), pp. 587–590 (2000)Google Scholar
- 10.Cardoso, J.-F., Souloumiac, A.: Blind beamforming for non Gaussian signals. IEE-Proceedings-F 140(6), 362–370 (1993)Google Scholar
- 14.Amari, S., Cichocki, A., Yang, H.H.: A new learning algorithm for blind signal separation. Advances in Neural Information Processing Systems 8, 757–763 (1996)Google Scholar