Dissertation > Excellent graduate degree dissertation topics show

The Research of Blind Source Separation Algorithm Based on Noisy Model

Author: ZhaoTianJiao
Tutor: HeXuanSen
School: Hunan University
Course: Information and Communication Engineering
Keywords: blind source separation independent component analysis FastICAalgorthm fifth-order convergence natural gradient algorithm waveletanalysis denoising
CLC: TN911.7
Type: Master's thesis
Year: 2011
Downloads: 68
Quote: 1
Read: Download Dissertation

Abstract


As the research hotspots in signal processing area, Blind source separation(BSS)has been successfully applied to the fields of image and the speech signal processing,biomedical signal analysis and processing, the antenna array signal processing and so on.The main purpose of BSS is only using the observation data received by the sensors toresume the source signals, while under the situation with unknown transmit channel andunknown sources.This paper has systematically explained the main solution of the blind sourceseparation problem----Independent Component Analysis (ICA). The basic assumptions,basic characters, optimization criterion,performances and evaluation index of ICAalgorithms are summarized. FastICA and natural gradient algorithms are analyzed basedon the model of instantaneous linear mixture blind source separation.Most blind source separation algorithms are established on noisy-free model.However, when the signal involves noise, the performance of these algorithms will dropeven defeat. In order to solve this problem, two kinds of noise separation model are setup, according to the different superposition of noise. On this basis, it summarizes fourmethods to solve noisy BBS. The classical BSS algorithms were improved to eliminatenoise caused by the deviation, such as Gaussian moments of FastICA algorithm and thebias removal technology of natural gradient algorithm. Wavelet analysis denoising isalso studied for BSS.The main innovations of this paper are proposed as follows:(1)FastICA algorithm is optimized by the quadratic convergence of Newtoniteration method and is widely applied in BBS. In order to accelerate the convergencespeed of the algorithm, this article gives two improved FastICA algorithms which use thefifth-order convergence of Newton iterative method. The experimental results show theperformance of the improved algorithms are better than the original algorithm.(2)A new wavelet threshold function is proposed by analyzing the main factorswhich affect the performance of wavelet threshold denoising.(3)The bias removal technology of natural gradient algorithm is used under thenoisy circumstances to estimate the demixing matrix. Then the new wavelet thresholdfunction is applied to separated signals to remove noise. This method is robust andimproves the SNR of the ultimate results obviously.Though BSS has matured to a point, there are still many problems and difficult in practical applications, so the direction of next research is briefly discussed at the end ofthis paper.

Related Dissertations

  1. Research on Noise Source Identification of Diesel Engine Based on Wavelet Analysis and Psychoacoustical Evaluation,TK421.6
  2. The Research of Information Hiding Algorithm Based on Noise Visibility Function and the Independent Component Analysis,TP391.41
  3. Paradigm Design and Algorithm Research for P300-based BCI,TP334.7
  4. Heart Sounds Digital Watermarking Processing Technology,TP309.7
  5. Research of Algorithms and Applications of Blind Signal Separation Under Underdetermined Condition,TN911.7
  6. Corrosion of the tank bottom Noise in AE Research,TH878
  7. Weak sparse underdetermined blind signal separation technology research,TN911.7
  8. Image feature extraction based image fusion research,TP391.41
  9. Inspection systems for metal abrasive noise cancellation algorithm,TP391.41
  10. Based on odor analysis equipment malfunction detection method,TB17
  11. Telephone-based channel voiceprint recognition algorithm,TN912.34
  12. Design of Visual Experimental Platform Based on Optical Imaging and Microelectrode Array,TP391.41
  13. Multi-Class Pattern Analysis on Human Brain MRI Dataset,TP391.41
  14. Study of Short-Term Power Forecasting in Coastal Wind Farm,TM614
  15. A Study of GPS Dynamic Deformation Monitoring Data Processing Based on EMD and ICA,P228.4
  16. Butadiene Rubber Plant and Its Performance Evaluation and Fault Localization Research,TQ333.2
  17. Study on Independent Component Analysis-Based Seismic Blind Deconvolution Method and Its Application,P631.4
  18. The Flaw Signal Extraction of Wire Rope Based on independent Component Analysis and Wavelet Transform,TG115.284
  19. Study on Repeated Blind Separation of Independent Component Analysis and Its Application in Mechanical Fault Diagnosis,TH165.3
  20. A Study of Lung Sound Extraction Based on Blind Source Separation,TN911.72
  21. Based on digital visual spelling - Listen combined stimulation ERP Study,R318.0

CLC: > Industrial Technology > Radio electronics, telecommunications technology > Communicate > Communication theory > Signal processing
© 2012 www.DissertationTopic.Net  Mobile