无线与移动通信中的信号处理新技术 英文版 第1册 信道估计与均衡PDF|Epub|txt|kindle电子书版本网盘下载

无线与移动通信中的信号处理新技术 英文版 第1册 信道估计与均衡PDF格式电子书版下载

注意：本站所有压缩包均有解压码： 点击下载压缩包解压工具

1 CHANNEL ESTIMATION AND EQUALIZATION USING HIGHER-ORDER STATISTICS1

1.1 Introduction1

1.2 Single-User Systems:Baud Rate Sampling4

1.2.1 Cumulant Matching4

1.2.2 Inverse Filter Criteria8

1.2.3 Equation Error Formulations8

1.2.4 Simulation Examples8

1.3 Single-User Systems:Fractional Sampling12

1.3.1 Cumulant Matching13

1.3.2 Simulation Example20

1.4 Multi-user Systems24

1.4.1 Inverse Filter Criteria26

1.4.2 Cumulant Matching28

1.4.3 Simulation Examples31

1.5 Concluding Remarks35

Bibliography37

2 PERFORMANCE BOUNDS FOR BLIND CHANNEL ESTIMATION41

2.1 Introduction42

2.2 Problem Statement and Preliminaries42

2.2.1 The Blind Channel Identification Problem42

2.2.2 Ambiguity Elimination44

2.2.3 The Unconstrained FIM46

2.2.4 Achievability of the CRB47

2.3 CRB for Constrained Estimates48

2.4 CRB for Estimates of Invariants49

2.5 CRB for Projection Errors52

2.6 Numerical Examples53

2.7 Concluding Remarks58

Appendix 2.A Proof of Proposition 259

Bibliography61

3 SUBSPACE METHOD FOR BLIND IDENTIFICATION AND DECONVOLUTION63

3.1 Introduction63

3.2 Subspace Identification of SIMO Channels65

3.2.1 Practical Considerations69

3.2.2 Simplifications in the Two-Channel Case70

3.3 Subspace Identification of MIMO Channels71

3.3.1 Rational Spaces and Polynomial Bases72

3.3.2 The Structure of the Left Nullspace of a Sylvester Matrix76

3.3.3 The Subspace Method78

3.3.4 Advanced Results82

3.4.1 Model Structure84

3.4 Applications to the Blind Channel Estimation of CDMA Systems84

3.4.2 The Structured Subspace Method:The Uplink Case88

3.4.3 The Structured Subspace Method:The Downlink Case89

3.5 Undermodeled Channel Identification92

3.5.1 Example:Identifying a Significant Part of a Channel99

3.5.2 Determining the Effective Impulse Response Length100

Appendix 3.A102

3.A.1 Proof of Theorem 1103

3.A.2 Proof of Proposition 3104

3.A.3 Proof of Theorem 4105

3.A.4 Proof of Proposition 5106

Bibliography108

4 BLIND IDENTIFICATION AND EQUALIZATION OF CHANNELS DRIVEN BY COLORED SIGNALS113

4.1 Introduction114

4.2.1 Original Model115

4.2.2 Slide-Window Formulation115

4.2 FIR MIMO Channel115

4.2.3 Noise Variance and Number of Input Signals116

4.3 Identifiability Using SOS117

4.3.1 Identifiability Conditions117

4.3.2 Some Facts of Polynomial Matrices118

4.3.3 Proof of the Conditions120

4.3.4 When the Input is White121

4.4 Blind Identification via Decorrelation121

4.4.1 The Principle of the BID121

4.4.2 Constructing the Decorrelators126

4.4.3 Removing the GCD of Polynomials128

4.4.4 Identification of the SIMO Channels130

Bibliography135

4.5 Final Remarks135

5 OPTIMUM SUBSPACE METHODS139

5.1 Introduction139

5.2 Data Model and Notations140

5.2.1 Scalar Valued Communication Systems140

5.2.2 Multi Channel Communication Systems141

5.2.3 A Stacked System Model143

5.2.4 Correlation Matrices145

5.2.5 Statistical Assumptions147

5.3 Subspace Ideas and Notations148

5.3.1 Basic Notations149

5.4 Parameterizations151

5.4.1 A Noise Subspace Parameterization151

5.4.2 Selection Matrices153

5.5 Estimation Procedure154

5.5.1 The Signal Subspace Parameterization155

5.5.2 The Noise Subspace Parameterization156

5.6 Statistical Analysis156

5.6.1 The Residual Covariance Matrices157

5.6.2 The Parameter Covariance Matrices159

5.7 Relation to Direction Estimation161

5.8 Further Results for the Noise Subspace Parameterization162

5.8.1 The Results163

5.8.2 The Approach163

5.9 Simulation Examples164

5.10 Conclusions171

Appendix 5.A173

Bibliography174

6 LINEAR PREDICTIVE ALGORITHMS FOR BLIND MULTICHANNEL IDENTIFICATION179

6.1 Introduction179

6.2 Channel Identification Based on Second Order Statistics:Problem Formulation181

6.3 Linear Prediction Algorithm for Channel Identification183

6.4 Outer-Product Decomposition Algorithm185

6.5 Multi-step Linear Prediction188

6.6 Channel Estimation by Linear Smoothing(Not Predicting)189

6.7 Channel Estimation by Constrained Output Energy Minirmization192

6.8 Discussion195

6.8.1 Channel Conditions195

6.8.2 Data Conditions196

6.8.3 Noise Effect196

6.9 Simulation Results197

6.10 Summary198

Bibliography207

7 SEMI-BLIND METHODS FOR FIR MULTICHANNEL ESTIMATION211

7.1.1 Training Sequence Based Methods and Blind Methods212

7.1 Introduction212

7.1.2 Semi-Blind Principle213

7.2 Problem Formulation214

7.3 Classification of Semi-Blind Methods217

7.4 Identifiability Conditions for Semi-Blind Channel Estimation218

7.4.1 Identifiability Definition218

7.4.2 TS Based Channel Identifiability219

7.4.3 Identifiability in the Deterministic Model219

7.4.4 Identifiability in the Gaussian Model222

7.5 Performance Measure:Cramér-Rao Bounds224

7.6 Performance Optimization Issues226

7.7 Optimal Semi-Blind Methods227

7.8 Blind DML229

7.8.1 Denoised IQML(DIQML)230

7.8.2 Pseudo Quadratic ML(PQML)231

7.9 Three Suboptimal DML Based Semi-Blind Criteria232

7.9.1 Split of the Data232

7.9.2 Least Squares-DML232

7.9.3 Alternating Quadratic DML(AQ-DML)233

7.9.4 Weighted-Least-Squares-PQML(WLS-PQML)235

7.9.5 Simulations236

7.10 Semi-Blind Criteria as a Combination of a Blind and a TS Based Criteria236

7.10.1 Semi-Blind SRM Example237

7.10.2 Subspace Fitting Example239

7.11 Performance of Semi-Blind Quadratic Criteria242

7.11.1 Mu and Mk infinite243

7.11.2 Mu infinite,Mk finite243

7.12 Gaussian Methods247

7.11.3 Optimally Weighted Quadratic Criteria247

7.13 Conclusion249

Bibliography250

8 A GEOMETRICAL APPROACH TO BLIND SIGNAL ESTIMATION255

8.1 Introduction256

8.2 Design Criteria for Blind Estimators258

8.2.1 The Constant Modulus Receiver260

8.2.2 The Shalvi-Weinstein Receiver261

8.3 The Signal Space Property and Equivalent Cost Functions263

8.3.1 The Signal Space Property of CM Receivers263

8.3.2 The Signal Space Property of SW Receivers264

8.3.3 Equivalent Cost Functions265

8.4 Geometrical Analysis of SW Receivers:Global Characterization266

8.4.1 The Noiseless Case268

8.4.2 The Noisy Case270

8.4.3 Domains of Attraction of SW Receivers275

8.5 Geometrical Analysis of SW Receivers:Local Characterizations277

8.5.1 Local Characterization277

8.5.2 MSE of CM Receivers281

8.6 Conclusion and Bibliography Notes282

8.6.1 Bibliography Notes283

Appendix 8.A Proof of Theorem 5285

Bibliography288

9 LINEAR PRECODING FOR ESTIMATION AND EQUALIZATION OF FREQUENCY-SELECTIVE CHANNELS291

9.1 System Model293

9.2 Unifying Filterbank Precoders296

9.3 FIR-ZF Equalizers301

9.4 Jointly Optimal Precoder and Decoder Design306

9.4.1 Zero-order Model306

9.4.2 MMSE/ZF Coding308

9.4.3 MMSE Solution with Constrained Average Power309

9.4.4 Constrained Power Maximum Information Rate Design311

9.4.5 Comparison Between Optimal Designs313

9.4.6 Asymptotic Performance317

9.4.7 Numerical Examples318

9.5 Blind Symbol Recovery320

9.5.1 Blind Channel Estimation322

9.5.2 Comparison with Other Blind Techniques324

9.5.3 Statistical Efficiency330

9.6 Conclusion332

Bibliography332

10 BLIND CHANNEL IDENTIFIABILITY WITH AN ARBITRARY LINEAR PRECODER339

10.1 Introduction339

10.2.2 General Properties of Polynomial Maps344

10.2 Basic Theory of Polynomial Equations344

10.2.1 Definition of Generic344

10.2.3 Generic and Non-Generic Points346

10.3 Inherent Scale Ambiguity348

10.4 Weak Identifiability and the CRB348

10.5 Arbitrary Linear Precoders349

10.6 Zero Prefix Precoders351

10.7 Geometric Interpretation of Precoding354

10.7.1 Linear Precoders354

10.7.2 Zero Prefix Precoders355

10.8 Filter Banks355

10.8.1 Algebraic Analysis of Filter Banks357

10.8.2 Spectral Analysis of Filter Banks358

10.9 Ambiguity Resistant Precoders360

10.10 Symbolic Methods361

10.11 Conclusion362

Bibliography363

11 CURRENT APPROACHES TO BLIND DECISION FEEDBACK EQU ALIZATION367

11.1 Introduction367

11.2 Notation370

11.3 Data Model373

11.4 Wiener Filtering374

11.4.1 Unconstrained Length MMSE Receivers375

11.4.2 Constrained Length MMSE Receivers377

11.4.3 Example:Constrained Versus Unconstrained Length Wiener Receivers379

11.5 Blind Tracking Algorithms380

11.5.1 DD-DFE381

11.5.2 CMA-DFE388

11.5.3 Algorithmic and Structural Modifications389

11.5.4 Summary of Blind Tracking Algorithms391

11.6 DFE Initialization Strategies391

11.6.1 Generic Strategy391

11.6.2 Multistage Equalization395

11.6.3 CMA-IIR Initialization397

11.6.4 Local Stability of Adaptive IIR Equalizers398

11.6.5 Summary of Blind Initialization Strategies399

11.7 Conclusion400

Appendix 11.A Spectral Factorization402

Appendix 11.B CL-MMSE-DFE403

Appendix 11.C DD-DFE Local Convergence405

Appendix 11.D Adaptive IIR Algorithm Updates406

Appendix 11.E CMA-AR Local Stability409

Bibliography411