Neural networks for RF and microwave design

Neural networks are information processing systems that can learn, generalize and even allow model development when component formulas are missing. This text demonstrates how to use neural networks to create models, evaluate models, and conquer some of the toughest RF and microwave CAD problems. The book also includes a review of neural netowrk applications in modelling microstrip lines, vias, CPW disconinuities, spiral inductors, FET and VLSI interconnects.

Preface	p. xv
1 Introduction and Overview	p. 1
1.1 RF and Microwave Design	p. 1
1.2 Artificial Neural Networks (ANNs)	p. 3
1.3 Overview of the Book	p. 4
References	p. 8
2 Modeling and Optimization for Design	p. 11
2.1 The Design Process	p. 11
2.2 RF and Microwave Circuit CAD	p. 19
2.3 CAD for Printed RF and Microwave Antennas	p. 38
2.4 Role of ANNs in RF and Microwave CAD	p. 55
2.5 Summary	p. 57
References	p. 57
3 Neural Network Structures	p. 61
3.1 Introduction	p. 61
3.2 Multilayer Perceptrons (MLP)	p. 64
3.3 Back Propagation (BP)	p. 75
3.4 Radial Basis Function Networks (RBF)	p. 77
3.5 Comparison of MLP and RBF Neural Networks	p. 81
3.6 Wavelet Neural Networks	p. 83
3.7 Arbitrary Structures	p. 88
3.8 Clustering Algorithms and Self-Organizing Maps	p. 90
3.9 Recurrent Neural Networks	p. 97
3.10 Summary	p. 100
References	p. 101
4 Training of Neural Networks	p. 105
4.1 Microwave Neural Modeling: Problem Statement	p. 105
4.2 Key Issues in Neural Model Development	p. 106
4.3 Neural Network Training	p. 128
4.4 Back Propagation Algorithm and Its Variants	p. 133
4.5 Training Algorithms Using Gradient-Based Optimization Techniques	p. 137
4.6 Nongradient-Based Training: Simplex Method	p. 141
4.7 Training With Global Optimization Methods	p. 143
4.8 Training Algorithms Utilizing Decomposed Optimization	p. 146
4.9 Comparisons of Different Training Techniques	p. 147
4.10 Feedforward Neural Network Training: Examples	p. 148
References	p. 151
5 Models for RF and Microwave Components	p. 155
5.1 Modeling Procedure	p. 155
5.2 Models for Vias and Multilayer Interconnects	p. 158
5.3 EM-ANN Models for CPW Components	p. 168
5.4 Other Passive Components' Models	p. 178
References	p. 190
6 Modeling of High-Speed IC Interconnects	p. 195
6.1 Introduction	p. 195
6.2 High-Speed Interconnect Modeling and Signal Integrity Analysis	p. 197
6.3 Application Examples	p. 203
6.4 Discussion	p. 216
6.5 Conclusions	p. 222
References	p. 223
7 Active Component Modeling Using Neural Networks	p. 227
7.1 Introduction	p. 227
7.2 Direct Modeling Approach	p. 228
7.3 Indirect Modeling Approach Through a Known Equivalent Circuit Model	p. 239
7.4 Discussion	p. 245
References	p. 246
8 Design Analysis and Optimization	p. 249
8.1 Design and Optimization Using ANN Models	p. 249
8.2 Optimization of Component Structure	p. 250
8.3 Circuit Optimization Using ANN Models	p. 251
8.4 Multilayer Circuit Design and Optimization Using ANN Models	p. 255
8.5 CPW Patch Antenna Design and Optimization	p. 265
8.6 Yield Optimization of a Three-Stage MMIC Amplifier	p. 278
8.7 Remarks	p. 278
References	p. 281
9 Knowledge-Based ANN Models	p. 283
9.1 Introduction	p. 283
9.2 Knowledge-Based Neural Networks (KBNN)	p. 285
9.3 Source Difference Method	p. 303
9.4 Prior Knowledge Input Method (PKI)	p. 310
9.5 Space-Mapped Neural Networks	p. 312
9.6 Hierarchical Neural Networks and Neural Model Library Development	p. 315
9.7 Summary	p. 333
References	p. 334
10 Concluding Remarks and Emerging Trends	p. 337
10.1 Summary of the Book	p. 337
10.2 Impact of Neural Nets on RF and Microwave Design	p. 340
10.3 Trends and Challenges	p. 342
References	p. 346
Appendix A NeuroModeler Introductory Version	p. 347
A.1 System Requirements	p. 347
A.2 How to Install the Software	p. 347
A.3 Quick Start the Program Using an Example	p. 348
A.4 User Interactions	p. 349
A.5 Highlights of the Introductory Version	p. 350
A.6 Information on Upgrade to the Standard Version	p. 350
About the Authors	p. 353
Index	p. 357

Available:*

On Order

Summary

Summary

Table of Contents