Cognitive Communications: Distributed Artificial Intelligence (DAI), Regulatory Policy & Economics, Implementation -- Contents -- List of Figures -- List of Tables -- About the Editors -- Preface -- PART I INTRODUCTION -- 1 Introduction to Cognitive Communications -- 1.1 Introduction -- 1.2 A NewWay of Thinking -- 1.3 History of Cognitive Communications -- 1.4 Key Components of Cognitive Communications -- 1.5 Overview of the Rest of the Book -- 1.5.1 Part 2: Wireless Communications -- 1.5.2 Part 3: Application of Distributed Artificial Intelligence -- 1.5.3 Part 4: Regulatory Policy and Economics -- 1.5.4 Part 5: Implementation -- 1.6 Summary and Conclusion -- References -- PART II WIRELESS COMMUNICATIONS -- 2 Cognitive Radio and Networks for Heterogeneous Networking -- 2.1 Introduction -- 2.1.1 Historical Sketch -- 2.1.2 Cognitive Radio and Networks -- 2.1.3 Heterogeneous Networks -- 2.2 Cognitive Radio for Heterogeneous Networks -- 2.2.1 Channel Sensing and Network Sensing -- 2.2.2 Interference Mitigation -- 2.2.3 Power Control -- 2.3 Applying Cognitive Networks to Heterogeneous Networks -- 2.3.1 Network Policy for Coexistence of Different Networks -- 2.3.2 Cooperation Mechanisms -- 2.3.3 Network Resource Allocation -- 2.3.4 Self-Organization Mechanisms -- 2.3.5 Handover Mechanisms -- 2.4 Performance Evaluation -- 2.5 Conclusion -- References -- 3 Channel Assignment and Power Allocation Algorithms in Multi-Carrier-Based Cognitive Radio Environments -- 3.1 Introduction -- 3.2 The Orthogonal Frequency-Division Multiplexing (OFDM) Transmission Scheme -- 3.2.1 Why OFDM is Appropriate for CR -- 3.3 Resource Management in Non-Cognitive OFDM Environments -- 3.3.1 Single User OFDM Systems -- 3.3.2 Multiple User OFDM Systems (OFDMA) -- 3.3.3 Resource Allocation Algorithms in Non-Cognitive OFDM Systems.

3.4 Resource Management in OFDM-Based Cognitive Radio Systems -- 3.4.1 Algorithms Dealing with In-Band Interference -- 3.4.2 Algorithms Dealing with Mutual Interference -- 3.4.3 System Model -- 3.4.4 Problem Formulation -- 3.4.5 Resource Management in Downlink OFDM-Based CR Systems -- 3.4.6 Resource Management in Uplink OFDM-Based CR Systems -- 3.5 Conclusions -- References -- 4 Filter Bank Techniques for Multi-Carrier Cognitive Radio Systems -- 4.1 Introduction -- 4.2 Basic Features of Filter Banks-Based Multi-Carrier Techniques -- 4.2.1 Introduction to the Filter Bank System -- 4.2.2 The Polyphase Structure of Filter Banks -- 4.2.3 Basic Structure of Filter Banks-Based Multi-Carrier Systems -- 4.3 Adaptive Threshold Enhanced Filter Bank for Spectrum Detection in IEEE 802.22 -- 4.3.1 Multi-Stage Analysis Filter Banks for Spectrum Detection -- 4.3.2 Complexity and Detection Precision Analysis -- 4.3.3 Spectrum Detection in IEEE 802.22 -- 4.3.4 Power Estimation with Adaptive Threshold -- 4.4 Transform Decomposition for Spectrum Interleaving in Multi-Carrier Cognitive Radio Systems -- 4.4.1 FFT Pruning in Cognitive Radio Systems -- 4.4.2 Transform Decomposition for General DFT -- 4.4.3 Improved Transform Decomposition Method for DFT with Sparse Input Points -- 4.4.4 Numerical Results and Computational Complexity Analysis -- 4.5 Remaining Problems in Filter Banks-Based Multi-Carrier Systems -- 4.6 Summary and Conclusion -- References -- 5 Distributed Clustering of Cognitive Radio Networks: A Message-Passing Approach -- 5.1 Introduction -- 5.1.1 Inter-Node Collaboration in Decentralized Cognitive Networks -- 5.1.2 Scalability Issues and Overhead Costs -- 5.1.3 Self-Organization Based on Distributed Clustering -- 5.2 Clustering Techniques for Cognitive Radio Networks -- 5.3 A Message-Passing Clustering Approach Based on Affinity Propagation.

5.4 Case Studies -- 5.4.1 Clustering Based on Local Spectrum Availability -- 5.4.2 Sensor Selection for Cooperative Spectrum Sensing -- 5.5 Implementation Challenges -- 5.6 Conclusions -- References -- PART III APPLICATION OF DISTRIBUTED ARTIFICIAL INTELLIGENCE -- 6 Machine Learning Applied to Cognitive Communications -- 6.1 Introduction -- 6.2 State of the Art -- 6.3 Learning Techniques -- 6.3.1 Bayesian Statistics -- 6.3.2 Supervised Neural Networks (NNs) -- 6.3.3 Self-Organizing Maps (SOMs): An Unsupervised Neural Network -- 6.3.4 Reinforcement Learning -- 6.4 Advantages and Disadvantages of Applying Machine Learning to Cognitive Radio Networks -- 6.5 Conclusions -- Acknowledgement -- References -- 7 Reinforcement Learning for Distributed Power Control and Channel Access in Cognitive Wireless Mesh Networks -- 7.1 Introduction -- 7.2 Applying Reinforcement Learning to Distributed Power Control and Channel Access -- 7.2.1 Conjecture-Based Multi-Agent Q-Learning for Distributed Power Control in CogMesh -- 7.2.2 Learning with Dynamic Conjectures for Opportunistic Spectrum Access in CogMesh -- 7.3 Future Challenges -- 7.4 Conclusions -- References -- 8 Reinforcement Learning-Based Cognitive Radio for Open Spectrum Access -- 8.1 Open Spectrum Access -- 8.2 Reinforcement Learning-Based Spectrum Sharing in Open Spectrum Bands -- 8.2.1 Learning Model -- 8.2.2 Basic Algorithms -- 8.2.3 Performance -- 8.3 Exploration Control and Efficient Exploration for Reinforcement Learning-Based Cognitive Radio -- 8.3.1 Exploration Control Techniques for Cognitive Radios -- 8.3.2 Efficient Exploration Techniques and Learning Efficiency for Cognitive Radios -- 8.4 Conclusion -- References -- 9 Learning Techniques for Context Diagnosis and Prediction in Cognitive Communications -- 9.1 Introduction -- 9.2 Prediction.

9.2.1 Building Knowledge: Learning Network Capabilities and User Preferences/ Behaviours -- 9.2.2 Application to Context Diagnosis and Prediction: The Case of Congestion -- 9.3 Future Problems -- 9.4 Conclusions -- References -- 10 Social Behaviour in Cognitive Radio -- 10.1 Introduction -- 10.2 Social Behaviour in Cognitive Radio -- 10.2.1 Cooperation Formation -- 10.2.2 Channel Recommendations -- 10.3 Social Network Analysis -- 10.3.1 Model of Recommendation Mechanism -- 10.3.2 Interacting Particles -- 10.3.3 Epidemic Propagation -- 10.4 Conclusions -- References -- PART IV REGULATORY POLICY AND ECONOMICS -- 11 Regulatory Policy and Economics of Cognitive Radio for Secondary Spectrum Access -- 11.1 Introduction -- 11.2 Spectrum Regulations: Why and How? -- 11.3 Overview of Regulatory Bodies and Their Inter-Relation -- 11.3.1 ITU -- 11.3.2 CEPT/ECC -- 11.3.3 European Union -- 11.3.4 ETSI -- 11.3.5 National Spectrum Management Authority -- 11.4 Why Secondary Spectrum Access? -- 11.5 Candidate Bands for Secondary Access -- 11.5.1 Terrestrial Broadcasting Bands -- 11.5.2 Radar Bands -- 11.5.3 IMT Bands -- 11.5.4 Military Bands -- 11.6 Regulatory and Policy Issues -- 11.6.1 UK Regulatory Environment -- 11.6.2 US Regulatory Environment -- 11.6.3 European Regulatory Environment -- 11.6.4 Regulatory Environments Elsewhere -- 11.7 Technology Enablers and Options for Secondary Sharing -- 11.7.1 Cognitive Radio -- 11.7.2 Technology Options for Secondary Access -- 11.8 Economic Impact and Business Opportunities of SSA -- 11.8.1 Stakeholders and Economic of SSA -- 11.8.2 Use Cases and Business Models -- 11.9 Outlook -- 11.10 Conclusions -- Acknowledgements -- References -- PART V IMPLEMENTATION -- 12 Cognitive Radio Networks in TV White Spaces -- 12.1 Introduction -- 12.2 Research and Development Challenges -- 12.2.1 Geolocation Databases -- 12.2.2 Sensing.

12.2.3 Beacons -- 12.2.4 Physical Layer -- 12.2.5 System Issues -- 12.2.6 Devices -- 12.3 Regulation and Standardization -- 12.3.1 Regulation -- 12.3.2 Standardization -- 12.4 Quantifying Spectrum Opportunities -- 12.5 Commercial Use Cases -- 12.6 Conclusions -- Acknowledgement -- References -- 13 Cognitive Femtocell Networks -- 13.1 Introduction -- 13.2 Femtocell Network Architecture -- 13.2.1 Underlay and Overlay Architectures for Femtocell Networks -- 13.2.2 Home Femtocell and Enterprise Femtocell -- 13.2.3 Access Mechanism: Closed, Open and Hybrid Access -- 13.2.4 Possible Operating Spectrum -- 13.3 Interference Management Strategies -- 13.3.1 Cross-Tier Interference Management -- 13.3.2 Intra-Tier Interference Management -- 13.4 Self Organized Femtocell Networks (SOFN) -- 13.4.1 Self-Configuration -- 13.4.2 Self-Optimization -- 13.4.3 Self-Healing and Self-Protection -- 13.5 Future Research Directions -- 13.5.1 Green Femtocell Networks -- 13.5.2 Communication Hub for Smart Homes -- 13.5.3 MIMO-Based Interference Alignment for Femtocell Networks -- 13.5.4 Enhanced FFR -- 13.5.5 CoMP-Based Femtocell Network -- 13.5.6 Holistic Approach to SOFN -- 13.6 Conclusion -- References -- 14 Cognitive Acoustics: A Way to Extend the Lifetime of Underwater Acoustic Sensor Networks -- 14.1 The Concept of Cognitive Acoustics -- 14.2 Underwater Acoustic Communication Channel -- 14.2.1 Propagation Delay -- 14.2.2 Severe Attenuation -- 14.2.3 Ambient Noise -- 14.3 Some Distinct Features of Cognitive Acoustics -- 14.3.1 Purposes of Deployment -- 14.3.2 Grey Space -- 14.3.3 Cost of Field Measurement and System Deployment -- 14.4 Fundamentals of Reinforcement Learning -- 14.4.1 Markov Decision Process -- 14.4.2 Reinforcement Learning -- 14.4.3 Q-Learning -- 14.5 An Application Scenario: Underwater Acoustic Sensor Networks -- 14.5.1 System Description.

14.5.2 State Space, Action Set and Transition Probabilities.