A new adaptive variable step size natural gradient BSS algorithm

2021 ◽  
pp. 1-12
Author(s):  
Junqing Ji ◽  
Xiaojia Kong ◽  
Yajing Zhang ◽  
Tongle Xu ◽  
Jing Zhang
Keyword(s):  
Signal To Noise Ratio ◽  
Wavelet Coefficient ◽  
Gradient Algorithm ◽  
Variable Step Size ◽  
Natural Gradient ◽  
Threshold Method ◽  
Step Size ◽  
Morphological Component Analysis ◽  
Variable Step ◽  
Separation Degree

The traditional blind source separation (BSS) algorithm is mainly used to deal with signal separation under the noiseless model, but it does not apply to data with the low signal to noise ratio (SNR). To solve the problem, an adaptive variable step size natural gradient BSS algorithm based on an improved wavelet threshold is proposed in this paper. Firstly, an improved wavelet threshold method is used to reduce the noise of the signal. Secondly, the wavelet coefficient layer with obvious periodicity is denoised using a morphological component analysis (MCA) algorithm, and the processed wavelet coefficients are recombined to obtain the ideal model. Thirdly, the recombined signal is pre-whitened, and a new separation matrix update formula of natural gradient algorithm is constructed by defining a new separation degree estimation function. Finally, the adaptive variable step size natural gradient blind source algorithm is used to separate the noise reduction signal. The results show that the algorithm can not only adaptively adjust the step size according to different signals, but also improve the convergence speed, stability and separation accuracy.

Study on the Noise Source Identification Based on a Novel Variable Step-Size Algorithm of Blind Sources Separation

2007 ◽  
Author(s):  
Linke Zhang ◽  
Lin He ◽  
Yong Jiang
Keyword(s):  
Output Feedback ◽  
Source Identification ◽  
Noise Source ◽  
Gradient Algorithm ◽  
Variable Step Size ◽  
Natural Gradient ◽  
Step Size ◽  
Input Signals ◽  
Noise Source Identification ◽  
Variable Step

Some preknowledge of sources input signals or transmission paths were required in advance for traditional noise source identification. In this paper, a novel variable step-size algorithm of blind source separation (BSS) is proposed to identify noise source, which doesn’t need any preknowledge but some statistical assumptions about sources. Most BSS algorithms have been issued on fixed step-size, relatively little work has been focused on variable step-size. The output feedback of step-size update in the proposed algorithm is derived from the analysis of adaptive blind sources separation and adaptive filter. With the natural gradient algorithm based minimum mutual information, the iteration formula of separate matrix is also obtained. The availability of this algorithm is confirmed by simulations. Namely, the convergence rate of this algorithm is rapid, while ensuring low steady-state error.

Research of Vector Hydrophone Adaptive Signal Processing

2013 ◽  
Vol 423-426 ◽  
pp. 2496-2506
Author(s):  
Bo Le Ma ◽  
Jing Fang Cheng ◽  
Wei Zhang
Keyword(s):  
Signal To Noise Ratio ◽  
Adaptive Signal Processing ◽  
Line Spectrum ◽  
Variable Step Size ◽  
Step Size ◽  
Spectrum Detection ◽  
Vector Hydrophone ◽  
Variable Step ◽  
Simulation And Experiment ◽  
Adaptive Signal

As a useful tool for line spectrum detection in underwater signal ,ALE has been used wildly. But there are still some problems to influence the effect of ALE. This paper gives three problems on ALE and analyses these.Then by the characteristics of vector hydrophone and a improved variable step size LMS,this paper constructs a cascade double input with variable step size based on vector hydrophone line spectrum enhancer . This algorithm restrains the noise of main channel twice ,meanwhile controls the noise in reference channel , so as to improve signal to noise ratio better. At the same time ,because of adopting the improved variable step size LMS, the steady-state error is reduced. From the results of simulation and experiment, the method presented in this paper can have a better effect of line spectrum enhancement.

Sign in / Sign up

Export Citation Format

Share Document