by Chi-Chiu Cheung. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1997. / Includes bibliographical references (leaves 124-131). / Abstract --- p.i / Acknowledgments --- p.iii / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- The Blind Signal Separation Problem --- p.1 / Chapter 1.2 --- Contributions of this Thesis --- p.3 / Chapter 1.3 --- Applications of the Problem --- p.4 / Chapter 1.4 --- Organization of the Thesis --- p.5 / Chapter 2 --- The Blind Signal Separation Problem --- p.7 / Chapter 2.1 --- The General Blind Signal Separation Problem --- p.7 / Chapter 2.2 --- Convolutive Linear Mixing Process --- p.8 / Chapter 2.3 --- Instantaneous Linear Mixing Process --- p.9 / Chapter 2.4 --- Problem Definition and Assumptions in this Thesis --- p.9 / Chapter 3 --- Literature Review --- p.13 / Chapter 3.1 --- Previous Works on Blind Signal Separation with Instantaneous Mixture --- p.13 / Chapter 3.1.1 --- Algebraic Approaches --- p.14 / Chapter 3.1.2 --- Neural approaches --- p.15 / Chapter 3.2 --- Previous Works on Blind Signal Separation with Convolutive Mixture --- p.20 / Chapter 4 --- The Information-theoretic ICA Scheme --- p.22 / Chapter 4.1 --- The Bayesian YING-YANG Learning Scheme --- p.22 / Chapter 4.2 --- The Information-theoretic ICA Scheme --- p.25 / Chapter 4.2.1 --- Derivation of the cost function from YING-YANG Machine --- p.25 / Chapter 4.2.2 --- Connections to previous information-theoretic approaches --- p.26 / Chapter 4.2.3 --- Derivation of the Algorithms --- p.27 / Chapter 4.2.4 --- Roles and Constraints on the Nonlinearities --- p.30 / Chapter 4.3 --- Direction and Motivation for the Analysis of the Nonlinearity --- p.30 / Chapter 5 --- Properties of the Cost Function and the Algorithms --- p.32 / Chapter 5.1 --- Lemmas and Corollaries --- p.32 / Chapter 5.1.1 --- Singularity of J(V) --- p.33 / Chapter 5.1.2 --- Continuity of J(V) --- p.34 / Chapter 5.1.3 --- Behavior of J(V) along a radially outward line --- p.35 / Chapter 5.1.4 --- Impossibility of divergence of the information-theoretic ICA al- gorithms with a large class of nonlinearities --- p.36 / Chapter 5.1.5 --- Number and stability of correct solutions in the 2-channel case --- p.37 / Chapter 5.1.6 --- Scale for the equilibrium points --- p.39 / Chapter 5.1.7 --- Absence of local maximum of J(V) --- p.43 / Chapter 6 --- The Algorithms with Cubic Nonlinearity --- p.44 / Chapter 6.1 --- The Cubic Nonlinearity --- p.44 / Chapter 6.2 --- Theoretical Results on the 2-Channel Case --- p.46 / Chapter 6.2.1 --- Equilibrium points --- p.46 / Chapter 6.2.2 --- Stability of the equilibrium points --- p.49 / Chapter 6.2.3 --- An alternative proof for the stability of the equilibrium points --- p.50 / Chapter 6.2.4 --- Convergence Analysis --- p.52 / Chapter 6.3 --- Experiments on the 2-Channel Case --- p.53 / Chapter 6.3.1 --- Experiments on two sub-Gaussian sources --- p.54 / Chapter 6.3.2 --- Experiments on two super-Gaussian sources --- p.55 / Chapter 6.3.3 --- Experiments on one super-Gaussian source and one sub-Gaussian source which are globally sub-Gaussian --- p.57 / Chapter 6.3.4 --- Experiments on one super-Gaussian source and one sub-Gaussian source which are globally super-Gaussian --- p.59 / Chapter 6.3.5 --- Experiments on asymmetric exponentially distributed signals .。 --- p.60 / Chapter 6.3.6 --- Demonstration on exactly and nearly singular initial points --- p.61 / Chapter 6.4 --- Theoretical Results on the 3-Channel Case --- p.63 / Chapter 6.4.1 --- Equilibrium points --- p.63 / Chapter 6.4.2 --- Stability --- p.66 / Chapter 6.5 --- Experiments on the 3-Channel Case --- p.66 / Chapter 6.5.1 --- Experiments on three pairwise globally sub-Gaussian sources --- p.67 / Chapter 6.5.2 --- Experiments on three sources consisting of globally sub-Gaussian and globally super-Gaussian pairs --- p.67 / Chapter 6.5.3 --- Experiments on three pairwise globally super-Gaussian sources --- p.69 / Chapter 7 --- Nonlinearity and Separation Capability --- p.71 / Chapter 7.1 --- Theoretical Argument --- p.71 / Chapter 7.1.1 --- Nonlinearities that strictly match the source distribution --- p.72 / Chapter 7.1.2 --- Nonlinearities that loosely match the source distribution --- p.72 / Chapter 7.2 --- Experiment Verification --- p.76 / Chapter 7.2.1 --- Experiments on reversed sigmoid --- p.76 / Chapter 7.2.2 --- Experiments on the cubic root nonlinearity --- p.77 / Chapter 7.2.3 --- Experimental verification of Theorem 2 --- p.77 / Chapter 7.2.4 --- Experiments on the MMI algorithm --- p.78 / Chapter 8 --- Implementation with Mixture of Densities --- p.80 / Chapter 8.1 --- Implementation of the Information-theoretic ICA scheme with Mixture of Densities --- p.80 / Chapter 8.1.1 --- The mixture of densities --- p.81 / Chapter 8.1.2 --- Derivation of the algorithms --- p.82 / Chapter 8.2 --- Experimental Verification on the Nonlinearity Adaptation --- p.84 / Chapter 8.2.1 --- Experiment 1: Two channels of sub-Gaussian sources --- p.84 / Chapter 8.2.2 --- Experiment 2: Two channels of super-Gaussian sources --- p.85 / Chapter 8.2.3 --- Experiment 3: Three channels of different signals --- p.89 / Chapter 8.3 --- Seeking the Simplest Workable Mixtures of Densities ......... .。 --- p.91 / Chapter 8.3.1 --- Number of components --- p.91 / Chapter 8.3.2 --- Mixture of two densities with only biases changeable --- p.93 / Chapter 9 --- ICA with Non-Kullback Cost Function --- p.97 / Chapter 9.1 --- Derivation of ICA Algorithms from Non-Kullback Separation Functionals --- p.97 / Chapter 9.1.1 --- Positive Convex Divergence --- p.97 / Chapter 9.1.2 --- Lp Divergence --- p.100 / Chapter 9.1.3 --- De-correlation Index --- p.102 / Chapter 9.2 --- Experiments on the ICA Algorithm Based on Positive Convex Divergence --- p.103 / Chapter 9.2.1 --- Experiments on the algorithm with fixed nonlinearities --- p.103 / Chapter 9.2.2 --- Experiments on the algorithm with mixture of densities --- p.106 / Chapter 10 --- Conclusions --- p.107 / Chapter A --- Proof for Stability of the Equilibrium Points of the Algorithm with Cubic Nonlinearity on Two Channels of Signals --- p.110 / Chapter A.1 --- Stability of Solution Group A --- p.110 / Chapter A.2 --- Stability of Solution Group B --- p.111 / Chapter B --- Proof for Stability of the Equilibrium Points of the Algorithm with Cubic Nonlinearity on Three Channels of Signals --- p.119 / Chapter C --- Proof for Theorem2 --- p.122 / Bibliography --- p.124

Algorithms for analyzing laser interferometry signals were developed and adopted to the computer based processing of small displacement measurements. These methods, matrix operation approach and fixed parameters approach, are based on signal phase calculation and are able to replace complex fringe counting electronic circuits. The matrix operation provides an approach for instantaneously displaying the results. The computer fixed parameters analysis allows the laser intensity to vary arbitrarily during a measurement. Displacement caused by a piezoelectric crystal was measured. Second order polynomial curve fitting was performed. The root mean square error is found to be 0.0086 μm in this 8-bit data acquisition system. CTEs of a fused silica plate and a tube were measured by an interferometry system. Signals were analyzed by both manual chart approach and computer based fixed parameters approach. Results agree well with published data. The accuracy of the CTE measurement system was 4 μ€, one third of the reference NBS SRM 739 suggested standard deviation. Out-of-plane and in-plane displacements can be measured independently from speckle interferometry. Their resolutions are 0.3164 μm/cycle for the out-of-plane configuration and 0.224 μm/cycle for the in-plane configuration with light incident angle of 45°. Optical systems with Fast Fourier Transform data analysis showed that the minimum detectable vibration amplitudes were 0.0065 μm, 0.0038 μm, and 0.0010 μm for the out-of-plane speckle, the in-plane speckle, and Michelson interferometry systems respectively. Resonance frequency of a steel rod was found by the optical non-contact sensing system. The modulus of elasticity calculated from the resonance frequency was close to the literature data, 182 GPa vs. 200 GPa. / Graduation date: 1992

Not available / text

No description available.

published_or_final_version / Electrical and Electronic Engineering / Master / Master of Philosophy

Wing-Keung Sim. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1999. / Includes bibliographical references (leaves 109-114). / Abstracts in English and Chinese. / Abstract --- p.2 / 摘要 --- p.3 / Acknowledgements --- p.4 / Contents --- p.5 / List of Figures --- p.8 / List of Tables --- p.9 / Introductions --- p.10 / Chapter 1.1 --- Motivation & Aims --- p.10 / Chapter 1.2 --- Contributions --- p.11 / Chapter 1.3 --- Structure of Thesis --- p.11 / Electrophysiological Spike Discrimination --- p.13 / Chapter 2.1 --- Introduction --- p.13 / Chapter 2.2 --- Cellular Physiology --- p.13 / Chapter 2.2.1 --- Action Potential --- p.13 / Chapter 2.2.2 --- Recording of Spikes Activities --- p.15 / Chapter 2.2.3 --- Demultiplexing of Multi-Neuron Recordings --- p.17 / Chapter 2.3 --- Application of Clustering for Mixed Spikes Train Separation --- p.17 / Chapter 2.3.1 --- Design Principles for Spike Discrimination Procedures --- p.17 / Chapter 2.3.2 --- Clustering Analysis --- p.18 / Chapter 2.3.3 --- Comparison of Clustering Techniques --- p.19 / Chapter 2.4 --- Literature Review --- p.19 / Chapter 2.4.1 --- Template Spike Matching --- p.19 / Chapter 2.4.2 --- Reduced Feature Matching --- p.20 / Chapter 2.4.3 --- Artificial Neural Networks --- p.21 / Chapter 2.4.4 --- Hardware Implementation --- p.21 / Chapter 2.5 --- Summary --- p.22 / Correlation of Perceived Headphone Sound Quality with Physical Parameters --- p.23 / Chapter 3.1 --- Introduction --- p.23 / Chapter 3.2 --- Sound Quality Evaluation --- p.23 / Chapter 3.3 --- Headphone Characterization --- p.26 / Chapter 3.3.1 --- Frequency Response --- p.26 / Chapter 3.3.2 --- Harmonic Distortion --- p.26 / Chapter 3.3.3 --- Voice-Coil Driver Parameters --- p.27 / Chapter 3.4 --- Statistical Correlation Measurement --- p.29 / Chapter 3.4.1 --- Correlation Coefficient --- p.29 / Chapter 3.4.2 --- t Test for Correlation Coefficients --- p.30 / Chapter 3.5 --- Summary --- p.31 / Algorithms --- p.32 / Chapter 4.1 --- Introduction --- p.32 / Chapter 4.2 --- Principal Component Analysis --- p.32 / Chapter 4.2.1 --- Dimensionality Reduction --- p.32 / Chapter 4.2.2 --- PCA Transformation --- p.33 / Chapter 4.2.3 --- PCA Implementation --- p.36 / Chapter 4.3 --- Traditional Clustering Methods --- p.37 / Chapter 4.3.1 --- Online Template Matching (TM) --- p.37 / Chapter 4.3.2 --- Online Template Matching Implementation --- p.40 / Chapter 4.3.3 --- K-Means Clustering --- p.41 / Chapter 4.3.4 --- K-Means Clustering Implementation --- p.44 / Chapter 4.4 --- Unsupervised Neural Learning --- p.45 / Chapter 4.4.1 --- Neural Network Basics --- p.45 / Chapter 4.4.2 --- Artificial Neural Network Model --- p.46 / Chapter 4.4.3 --- Simple Competitive Learning (SCL) --- p.47 / Chapter 4.4.4 --- SCL Implementation --- p.49 / Chapter 4.4.5 --- Adaptive Resonance Theory Network (ART). --- p.50 / Chapter 4.4.6 --- ART2 Implementation --- p.53 / Chapter 4.6 --- Summary --- p.55 / Experimental Design --- p.57 / Chapter 5.1 --- Introduction --- p.57 / Chapter 5.2 --- Electrophysiological Spike Discrimination --- p.57 / Chapter 5.2.1 --- Experimental Design --- p.57 / Chapter 5.2.2 --- Extracellular Recordings --- p.58 / Chapter 5.2.3 --- PCA Feature Extraction --- p.59 / Chapter 5.2.4 --- Clustering Analysis --- p.59 / Chapter 5.3 --- Correlation of Headphone Sound Quality with physical Parameters --- p.61 / Chapter 5.3.1 --- Experimental Design --- p.61 / Chapter 5.3.2 --- Frequency Response Clustering --- p.62 / Chapter 5.3.3 --- Additional Parameters Measurement --- p.68 / Chapter 5.3.4 --- Listening Tests --- p.68 / Chapter 5.3.5 --- Confirmation Test --- p.69 / Chapter 5.4 --- Summary --- p.70 / Results --- p.71 / Chapter 6.1 --- Introduction --- p.71 / Chapter 6.2 --- Electrophysiological Spike Discrimination: A Comparison of Methods --- p.71 / Chapter 6.2.1 --- Clustering Labeled Spike Data --- p.72 / Chapter 6.2.2 --- Clustering of Unlabeled Data --- p.78 / Chapter 6.2.3 --- Remarks --- p.84 / Chapter 6.3 --- Headphone Sound Quality Control --- p.89 / Chapter 6.3.1 --- Headphones Frequency Response Clustering --- p.89 / Chapter 6.3.2 --- Listening Tests --- p.90 / Chapter 6.3.3 --- Correlation with Measured Parameters --- p.90 / Chapter 6.3.4 --- Confirmation Listening Test --- p.92 / Chapter 6.4 --- Summary --- p.93 / Conclusions --- p.97 / Chapter 7.1 --- Future Work --- p.98 / Chapter 7.1.1 --- Clustering Analysis --- p.98 / Chapter 7.1.2 --- Potential Applications of Clustering Analysis --- p.99 / Chapter 7.2 --- Closing Remarks --- p.100 / Appendix --- p.101 / Chapter A.1 --- Tables of Experimental Results: (Spike Discrimination) --- p.101 / Chapter A.2 --- Tables of Experimental Results: (Headphones Measurement) --- p.104 / Bibliography --- p.109 / Publications --- p.114

by Wing-Kit Chung. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2000. / Includes bibliographical references (leaves 123-127). / Abstracts in English and Chinese. / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Application of Time Delay and Doppler Shift Estimation in Active Radar --- p.1 / Chapter 1.2 --- Joint Time delay and Doppler Shift Estimation Algorithm based on Correlation --- p.4 / Chapter 1.3 --- A Brief Review of Radar Signal Design --- p.6 / Chapter 1.3.1 --- Suppression of Range Sidelobes Level --- p.6 / Chapter 1.3.2 --- Reduction of Ambiguity of Delay-Doppler Plane --- p.8 / Chapter 1.4 --- Goal and Outline of the Thesis --- p.9 / Chapter 2 --- CAF and Pulse Diversity for Radar Signals --- p.11 / Chapter 2.1 --- Radar Ambiguity Function --- p.12 / Chapter 2.1.1 --- Properties of Radar Ambiguity Function --- p.12 / Chapter 2.1.2 --- Ideal Ambiguity Function --- p.13 / Chapter 2.2 --- Composite Ambiguity Function (CAF) --- p.14 / Chapter 2.2.1 --- Properties of Composite Ambiguity Function --- p.15 / Chapter 2.3 --- CAF of Joint Phase and Frequency Shift Keying (PSK-FSK) Mod- ulated Signal --- p.17 / Chapter 2.4 --- Summary --- p.21 / Chapter 3 --- CAF Algorithm and Its Performance Analysis --- p.22 / Chapter 3.1 --- CAF Algorithm for Time Delay and Doppler Shift Estimation --- p.23 / Chapter 3.2 --- The Cramer-Rao Lower Bound of the CAF Algorithm --- p.24 / Chapter 3.3 --- Performance Analysis of the CAF Algorithm --- p.28 / Chapter 3.4 --- Global Accuracy --- p.31 / Chapter 3.5 --- Numerical Results for Derivation of CAF Algorithm --- p.35 / Chapter 3.5.1 --- Simulation Results of CRLB for Various Multi-pulse Signals --- p.35 / Chapter 3.5.2 --- Simulation Results of Global Accuracy for Various Multi- pulse Signals --- p.36 / Chapter 3.5.3 --- Simulation on Global Accuracy with Different Parameters --- p.37 / Chapter 3.6 --- Summary --- p.39 / Chapter 4 --- Optimum Pulse-Diverse Waveforms Design --- p.46 / Chapter 4.1 --- Criteria for Optimum Waveforms --- p.46 / Chapter 4.2 --- Optimum Signals Based on Joint Phase and Frequency Shift Key- ing (PSK-FSK) Modulated Signal --- p.48 / Chapter 4.3 --- Genetic Algorithm (GA) --- p.50 / Chapter 4.4 --- Numerical Results --- p.54 / Chapter 4.4.1 --- "Comparison of Optimized PSK, FSK and PSK-FSK Signals" --- p.55 / Chapter 4.4.2 --- Simulation on Large Number of Pulses for Pulse-diverse Waveform Set --- p.59 / Chapter 4.4.3 --- Simulation Results of CAF algorithm for Time Delay and Doppler Shift Estimation --- p.63 / Chapter 4.4.4 --- Various Distribution of Ambiguity Volume on the Delay- Doppler Plane --- p.70 / Chapter 4.5 --- Summary --- p.74 / Chapter 5 --- Wideband CAF (WCAF) and Its Analysis --- p.75 / Chapter 5.1 --- WCAF Algorithm for Time Delay and Doppler Stretch Estimation --- p.76 / Chapter 5.2 --- Theory of Wavelet Packets --- p.77 / Chapter 5.3 --- Design of Wideband Optimum Waveforms for WCAF Algorithm --- p.80 / Chapter 5.4 --- Performance Evaluation --- p.82 / Chapter 5.4.1 --- The Cramer-Rao Lower Bound of WCAF Algorithm --- p.83 / Chapter 5.4.2 --- The Global Accuracy of WCAF Algorithm --- p.84 / Chapter 5.4.3 --- Numerical Results --- p.86 / Chapter 5.5 --- Summary --- p.89 / Chapter 6 --- Conclusion and Suggestion for Future Research --- p.90 / Chapter 6.1 --- Conclusion --- p.90 / Chapter 6.2 --- Suggestion for Future Research --- p.93 / Chapter A --- Derivation of Ambiguity Function and CAF --- p.94 / Chapter A.1 --- Properties of Radar Ambiguity Function --- p.94 / Chapter A.2 --- Properties of Composite Ambiguity Function --- p.96 / Chapter B --- Derivation of Fisher Information Matrix of CAF Algorithm --- p.98 / Chapter C --- Derivation of Performance Analysis of CAF Algorithm --- p.103 / Chapter C.1 --- Derivation of TD and DS Estimate by Proposed Estimator --- p.103 / Chapter C.2 --- Derivation the Asymptotic Variance of The Estimates --- p.106 / Chapter D --- Derivation of Probability of Decision Error --- p.113 / Chapter E --- PSK-FSK Modulating Code of Various Multi-pulse Signals --- p.116 / Chapter F --- Derivation of Wavelet-Based Wideband CAF --- p.120 / Chapter F.1 --- Volume of Wideband Ambiguity Function --- p.120 / Chapter F.2 --- Volume of Wideband Composite Ambiguity Function --- p.121 / Bibliography --- p.123

由於在盲語音分離的應用，線性準平穩源訊號混合的盲識別獲得了巨大的研究興趣。在這個問題上，我們利用準穩態源訊號的時變特性來識別未知的混合系統系數。傳統的方法有二：i)基於張量分解的平行因子分析(PARAFAC);ii)基於對多個矩陣的聯合對角化的聯合對角化算法(JD)。一般來說，PARAFAC和JD 都採用了源聯合的提取方法；即是說，對應所有訊號源的系統係數在升法上是用時進行識別的。 / 在這篇論文中，我利用Khati-Rao(KR)子空間來設計一種新的盲識別算法。在我設計的算法中提出一種與傳統的方法不同的提法。在我設計的算法中，盲識別問題被分解成數個結構上相對簡單的子問題，分別對應不同的源。在超定混合模型，我們提出了一個專門的交替投影算法(AP)。由此產生的算法，不但能從經驗發現是非常有競爭力的，而且更有理論上的利落收斂保證。另外，作為一個有趣的延伸，該算法可循一個簡單的方式應用於欠混合模型。對於欠定混合模型，我們提出啟發式的秩最小化算法從而提高算法的速度。 / Blind identification of linear instantaneous mixtures of quasi-stationary sources (BI-QSS) has received great research interest over the past few decades, motivated by its application in blind speech separation. In this problem, we identify the unknown mixing system coefcients by exploiting the time-varying characteristics of quasi-stationary sources. Traditional BI-QSS methods fall into two main categories: i) Parallel Factor Analysis (PARAFAC), which is based on tensor decomposition; ii) Joint Diagonalization (JD), which is based on approximate joint diagonalization of multiple matrices. In both PARAFAC and JD, the joint-source formulation is used in general; i.e., the algorithms are designed to identify the whole mixing system simultaneously. / In this thesis, I devise a novel blind identification framework using a Khatri-Rao (KR) subspace formulation. The proposed formulation is different from the traditional formulations in that it decomposes the blind identication problem into a number of per-source, structurally less complex subproblems. For the over determined mixing models, a specialized alternating projections algorithm is proposed for the KR subspace for¬mulation. The resulting algorithm is not only empirically found to be very competitive, but also has a theoretically neat convergence guarantee. Even better, the proposed algorithm can be applied to the underdetermined mixing models in a straightforward manner. Rank minimization heuristics are proposed to speed up the algorithm for the underdetermined mixing model. The advantages on employing the rank minimization heuristics are demonstrated by simulations. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Lee, Ka Kit. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2012. / Includes bibliographical references (leaves 72-76). / Abstracts also in Chinese. / Abstract --- p.i / Acknowledgement --- p.ii / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- Settings of Quasi-Stationary Signals based Blind Identification --- p.4 / Chapter 2.1 --- Signal Model --- p.4 / Chapter 2.2 --- Assumptions --- p.5 / Chapter 2.3 --- Local Covariance Model --- p.7 / Chapter 2.4 --- Noise Covariance Removal --- p.8 / Chapter 2.5 --- Prewhitening --- p.9 / Chapter 2.6 --- Summary --- p.10 / Chapter 3 --- Review on Some Existing BI-QSS Algorithms --- p.11 / Chapter 3.1 --- Joint Diagonalization --- p.11 / Chapter 3.1.1 --- Fast Frobenius Diagonalization [4] --- p.12 / Chapter 3.1.2 --- Pham’s JD [5, 6] --- p.14 / Chapter 3.2 --- Parallel Factor Analysis --- p.16 / Chapter 3.2.1 --- Tensor Decomposition [37] --- p.17 / Chapter 3.2.2 --- Alternating-Columns Diagonal-Centers [12] --- p.21 / Chapter 3.2.3 --- Trilinear Alternating Least-Squares [10, 11] --- p.23 / Chapter 3.3 --- Summary --- p.25 / Chapter 4 --- Proposed Algorithms --- p.26 / Chapter 4.1 --- KR Subspace Criterion --- p.27 / Chapter 4.2 --- Blind Identification using Alternating Projections --- p.29 / Chapter 4.2.1 --- All-Columns Identification --- p.31 / Chapter 4.3 --- Overdetermined Mixing Models (N > K): Prewhitened Alternating Projection Algorithm (PAPA) --- p.32 / Chapter 4.4 --- Underdetermined Mixing Models (N <K) --- p.34 / Chapter 4.4.1 --- Rank Minimization Heuristic --- p.34 / Chapter 4.4.2 --- Alternating Projections Algorithm with Huber Function Regularization --- p.37 / Chapter 4.5 --- Robust KR Subspace Extraction --- p.40 / Chapter 4.6 --- Summary --- p.44 / Chapter 5 --- Simulation Results --- p.47 / Chapter 5.1 --- General Settings --- p.47 / Chapter 5.2 --- Overdetermined Mixing Models --- p.49 / Chapter 5.2.1 --- Simulation 1 - Performance w.r.t. SNR --- p.49 / Chapter 5.2.2 --- Simulation 2 - Performance w.r.t. the Number of Available Frames M --- p.49 / Chapter 5.2.3 --- Simulation 3 - Performance w.r.t. the Number of Sources K --- p.50 / Chapter 5.3 --- Underdetermined Mixing Models --- p.52 / Chapter 5.3.1 --- Simulation 1 - Success Rate of KR Huber --- p.53 / Chapter 5.3.2 --- Simulation 2 - Performance w.r.t. SNR --- p.54 / Chapter 5.3.3 --- Simulation 3 - Performance w.r.t. M --- p.54 / Chapter 5.3.4 --- Simulation 4 - Performance w.r.t. N --- p.56 / Chapter 5.4 --- Summary --- p.56 / Chapter 6 --- Conclusion and Future Works --- p.58 / Chapter A --- Convolutive Mixing Model --- p.60 / Chapter B --- Proofs --- p.63 / Chapter B.1 --- Proof of Theorem 4.1 --- p.63 / Chapter B.2 --- Proof of Theorem 4.2 --- p.65 / Chapter B.3 --- Proof of Observation 4.1 --- p.65 / Chapter B.4 --- Proof of Proposition 4.1 --- p.66 / Chapter C --- Singular Value Thresholding --- p.67 / Chapter D --- Categories of Speech Sounds and Their Impact on SOSs-based BI-QSS Algorithms --- p.69 / Chapter D.1 --- Vowels --- p.69 / Chapter D.2 --- Consonants --- p.69 / Chapter D.1 --- Silent Pauses --- p.70 / Bibliography --- p.72

by Xin-xin Niu. / Thesis (Ph.D.)--Chinese University of Hong Kong, 1997. / Includes bibliographical references. / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Mode of access: World Wide Web.

Zheng Ming. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2003. / Includes bibliographical references (leaves 63-66). / Abstracts in English and Chinese. / Chapter 1. --- Introduction --- p.1 / Chapter 1.1 --- Review of space-time coding and blind channel estimation --- p.1 / Chapter 1.2 --- Introduction of space-time coding system --- p.4 / Chapter 1.3 --- Diversity gain of space-time coding --- p.6 / Chapter 1.4 --- Re-estimation --- p.7 / Chapter 1.5 --- Notations --- p.8 / Chapter 1.6 --- Outline of thesis --- p.8 / Chapter 2. --- Estimation for BPSK Signals --- p.10 / Chapter 2.1 --- Introduction to maximum likelihood estimation --- p.10 / Chapter 2.2 --- System model --- p.11 / Chapter 2.3 --- Deterministic ML algorithm --- p.14 / Chapter 2.4 --- Re-estimation --- p.16 / Chapter 2.5 --- Application to other constellations --- p.18 / Chapter 2.6 --- Simulation results --- p.18 / Chapter 2.7 --- Summary --- p.21 / Chapter 3. --- Estimation for Flat Fading Channels --- p.22 / Chapter 3.1 --- Introduction of constant modulus algorithm (CMA) --- p.22 / Chapter 3.2 --- System model for flat fading channels --- p.24 / Chapter 3.3 --- Blind estimation with CMA --- p.26 / Chapter 3.3.1 --- Problem statement --- p.26 / Chapter 3.3.2 --- Estimating channel with CMA --- p.28 / Chapter 3.3.3 --- Solving the ambiguity problem --- p.32 / Chapter 3.4 --- Re-estimation for flat fading channels --- p.39 / Chapter 3.5 --- Estimation algorithm --- p.39 / Chapter 3.6 --- Application to multi-antenna system --- p.41 / Chapter 3.7 --- Simulation results --- p.42 / Chapter 3.8 --- Summary --- p.46 / Chapter 4. --- Estimation lor Frequency Selective Fading Channels --- p.48 / Chapter 4.1 --- Introduction of space-time coded OFDM --- p.48 / Chapter 4.2 --- System model --- p.51 / Chapter 4.3 --- Estimation Algorithm --- p.54 / Chapter 4.4 --- Simulation results --- p.56 / Chapter 4.5 --- Summary --- p.59 / Chapter 5. --- Conclus ions and Future Work --- p.60 / Chapter 5.1 --- Conclusions --- p.60 / Chapter 5.2 --- Future work --- p.61 / Bibliography: --- p.63