臺灣博碩士論文加值系統

在真實世界的多感應器應用中，如何從混合訊號中分析出獨立訊號的瞎訊號分離是一個常見的問題，例如:音訊和生醫訊號處理。本論文提出一個基於資訊最大化之獨立事件分析方法應用於四通道EEG訊號分離。並用定點數實現於FPGA,再藉由藍芽傳輸分離後的訊號。經由實驗的結果，本論文所提出的硬體方式比軟體運算快56倍，且絕對相關係數和離線訊號處理比較至少有80% 。 最後，實際示範將用Altera DE2發展板展示，此設計使用16605邏輯單元。
而本論文所提出的四通道即時獨立事件分析系統也加入彈性的介面用於實際EEG訊號分離的應用。用資訊最大化演算法的即時生醫訊號分離其取樣頻率設定在64Hz，並藉由整合性的算術運算架構可讓整體操作速度在68MHz。

Blind source separation of independent sources from their mixtures is a common problem for multi-sensor applications in real world, for example, speech or biomedical signal processing. This thesis presents an independent component analysis (ICA) method with information maximization (Infomax) update applied into 4-channel one-line EEG signal separation. This can be implemented on FPGA with a fixed-point number representation, and then the separated signals are transmitted via Bluetooth. As experimental results, the proposed design is faster 56 times than soft performance, and the correlation coefficients at least 80% with the absolute value are compared with off-line processing results. Finally, live demonstration is shown in the DE2 FPGA board, and the design is consisted of 16,605 logic elements.
The 4-channel On-line ICA accompanied with flexible communication interface for real EEG signal separation has been presented in this thesis. The proposed integrated mathematics architecture can allow high-speed at 68MHz and real-time biomedical signal separation with Infomax ICA at sampling rate 64 Hz.

中文摘要 ii
Abstract iii
誌謝 iv
List of Figures vii
List of Tables x
Chapter 1 Introduction 1
1-1 Motivation 1
1-2 Goal and Summary 2
1-3 Organization of the Thesis 2
Chapter 2 ICA Algorithm 3
2-1 Basic Concepts of ICA 3
2-1-1 Problem Description 3
2-1-2 Formulation 4
2-1-3 Independent Conditions 5
2-2 Two kinds of ICA Algorithm 6
2-2-1 The Concept of Entropy and Mutual Information 6
2-2-2 Infomax ICA 8
2-2-3 FAST ICA 13
2-3 Main Structure of ICA Methods 18
2-3-1 The Choice of ICA Algorithm 18
2-3-2 Solution of on-line ICA 18
Chapter 3 System Architecture Design and Simulation Result 21
3-1 System Architecture 21
3-1-1 Computing Flow 21
3-1-2 Specification of On-line Process 23
3-2 Comparison with off-line Sup-Gaussian BSS Methods 24
3-2-1 Simulation 8-bit Super Gaussian Mixed Pattern 1 24
3-2-2 Simulation 8-bit Super Gaussian Mixed Pattern 2 28
3-3 Comparison with off-line EEG BSS Methods 31
3-3-1 Simulation 8-bit EEG Mixed Pattern 1 31
3-3-2 Simulation 8-bit EEG Mixed Pattern 2 34
3-4 Summery of Comparison 37
Chapter 4 Implementation of the On-line ICA System on FPGA 38
4-1 Architecture of real-time Systems 38
4-1-1 Implementation of Recursive Operation Circuit 39
4-1-2 Implementation of system Controller 46
4-1-3 Interface Design 51
4-2 FPGA Simulation Result in Integrated System 55
4-2-1 FPGA Simulation in Recursive Operation Circuit 55
4-2-2 FPGA Simulation in System Controller 57
4-2-3 FPGA Simulation in Interface Design 59
4-2-4 FPGA Simulation in Integrated System 60
4-3 Device for Demonstration 62
4-3-1 Four channel EEG Brain-computer Interface 63
4-3-2 Wireless Transmission Model 67
4-3-3 GUI for Display 69
4-4 Summary 70
Chapter 5 Experimental Results 71
5-1 Result Super Gaussian BSS Methods in GUI 71
5-2 Comparison with other ICA Design 75
Chapter 6 Conclusion 77
References 79

[1]T-W Lee, "Independent Component Analysis - Theory and Applications", Kluwer Academic Publishers, 1998.

[2]C. M. Kim and S. Y. Lee, “A digital chip for robust speech recognition in noisy environment,” in Proc. IEEE Int. Conf. Acoustics, Speech, and Signal Processing, Vol. 2, pp. 1089–1092, 2001.

[3]Saruwatari, H., Kawamura, T., Sawai, K.; Kaminuma, A., Sakata, M, “Blind source separation based on fast-convergence algorithm using ICA and beamforming for real convolutive mixture,” IEEE International Conference on Acoustics, Speech, and Signal Processing, Vol. 1, pp. 13-17, 2002

[4]Ristaniemi, T.and Joutsensalo, J. “Advanced ICA-based receivers for DS-CDMA systems,” Personal, Indoor and Mobile Radio Communications, Vol. 1, pp. 276 -281, 2000

[5]A. J. Bell and T. J. Sejnowski, “An information maximization approach to blind separation and blind deconvolution,” Neurocomputing, Vol. 7, pp. 1129- 1159, 1995.

[6]K. Torkkola, "Blind separation of convolved sources based on information maximization", IEEE Workshop Neural Networks for Signal Processing, Kyoto, Japan, Sept 4-6, 1996.

[7]A. Hyvärinen and E. Oja. “A Fast Fixed-Point Algorithm for Independent Component Analysis,” Neural Computation, Vol. 9, pp. 1483-1492, 1997

[8]A. Hyvärinen. “Fast and Robust Fixed-Point Algorithms for Independent Component Analysis.” IEEE Transactions on Neural Networks 10(3), 626-634, 1999.

[9]C. Jutten and J. Herault, “Blind Separation of Sources 1. An Adaptive Algorithm Based on Neuromimetic Architecture,” Signal Processing, Vol. 24, pp 1-10, 1991.

[10]J.F. Cardoso, A. Souloumiac, “Blind Beamforming for Non-Gaussian Signals,” IEE Proc.F Vol. 140, pp. 362-370, 1993.

[11]A. Hyvärinen and E. Oja. “A Fast Fixed-Point Algorithm for Independent Component Analysis,” Neural Computation, Vol. 9, pp. 1483-1492, 1997

[12]C. Charoensak, and F. Sattar, "System-level design of low-cost FPGA hardware for real-time ICA-based blind source separation," SOC Con. 2004. Proceedings. IEEEInternat ional, pp. 139 - 140, 2004.

[13]H. Amin., K.M. Curtis, and B.R. Hayes-Gill, “Piecewise Linear Approximation Applied to Nonlinear Function of a Neural Network, ” IEE Proc. Crcuits Divices syst, Vol. 144, pp. 313-3171, 1997

[14]F. Sattar and C. Charayaphan, Low-cost design and implementation of an ICA-based blind source separation algorithm, 15th Annual IEEE International ASIC/SOC Conference, pp.15-19, 2002.

[15]H. Du and H. Qi, “An FPGA implementation of parallel ICA for dimensionality reduction in hyperspectral images,” in Proc. IEEE Int. Symp. Geosci. Remote Sens., Sep. 2004, vol. 5, pp. 3257–3260.

[16]C. Charoensak and F. Sattar, “A single-chip FPGA design for real-time ICA-based blind source separation algorithm,” in Proc. IEEE Int. Symp. Circuits Syst., Vol. 6, pp. 5822–5825, 2005.

[17]Kuo-Kai Shyu and Ming-Huan Lee, “Implementation of Pipelined FastICA on FPGA for Real-Time Blind Source Separation” in Proc IEEE TRANSACTIONS ON NEURAL NETWORKS, vol. 19, 2008