Wireless Sensor Networks -- Contents -- List of Contributors -- 1 Introduction -- Part I Fundamental Properties and Limits -- 2 Information-theoretic Bounds on Sensor Network Performance -- 2.1 Introduction -- 2.2 Sensor Network Models -- 2.2.1 The Linear Gaussian Sensor Network -- 2.3 Digital Architectures -- 2.3.1 Distributed Source Coding -- 2.3.2 Distributed Channel Coding -- 2.3.3 End-to-end Performance of Digital Architectures -- 2.4 The Price of Digital Architectures -- 2.5 Bounds on General Architectures -- 2.6 Concluding Remarks -- Bibliography -- 3 In-Network Information Processing in Wireless Sensor Networks -- 3.1 Introduction -- 3.2 Communication Complexity Model -- 3.3 Computing Functions over Wireless Networks: Spatial Reuse and Block Computation -- 3.3.1 Geographical Models of Wireless Communication Networks -- 3.3.2 Block Computation and Computational Throughput -- 3.3.3 Symmetric Functions and Types -- 3.3.4 The Collocated Network -- 3.3.5 Subclasses of Symmetric Functions: Type-sensitive and Type-threshold -- 3.3.6 Results on Maximum Throughput in Collocated Networks -- 3.3.7 Multi-Hop Networks: The Random Planar Network -- 3.3.8 Other Acyclic Networks -- 3.4 Wireless Networks with Noisy Communications: Reliable Computation in a Collocated Broadcast Network -- 3.4.1 The Sum of the Parity of the Measurements -- 3.4.2 Threshold Functions -- 3.5 Towards an Information Theoretic Formulation -- 3.6 Conclusion -- Bibliography -- 4 The Sensing Capacity of Sensor Networks -- 4.1 Introduction -- 4.1.1 Large-Scale Detection Applications -- 4.1.2 Sensor Network as an Encoder -- 4.1.3 Information Theory Context -- 4.2 Sensing Capacity of Sensor Networks -- 4.2.1 Sensor Network Model with Arbitrary Connections -- 4.2.2 Random Coding and Method of Types -- 4.2.3 Sensing Capacity Theorem -- 4.2.4 Illustration of Sensing Capacity Bound.

4.3 Extensions to Other Sensor Network Models -- 4.3.1 Models with Localized Sensing -- 4.3.2 Target Models -- 4.4 Conclusion -- Bibliography -- 5 Law of Sensor Network Lifetime and Its Applications -- 5.1 Introduction -- 5.2 Law of Network Lifetime and General Design Principle -- 5.2.1 Network Characteristics and Lifetime Definition -- 5.2.2 Law of Lifetime -- 5.2.3 A General Design Principle For Lifetime Maximization -- 5.3 Fundamental Performance Limit: A Stochastic Shortest Path Framework -- 5.3.1 Problem Statement -- 5.3.2 SSP Formulation -- 5.3.3 Fundamental Performance Limit on Network Lifetime -- 5.3.4 Computing the Limiting Performance with Polynomial Complexity in Network Size -- 5.4 Distributed Asymptotically Optimal Transmission Scheduling -- 5.4.1 Dynamic Protocol for Lifetime Maximization -- 5.4.2 Dynamic Nature of DPLM -- 5.4.3 Asymptotic Optimality of DPLM -- 5.4.4 Distributed Implementation -- 5.4.5 Simulation Studies -- 5.5 A Brief Overview of Network Lifetime Analysis -- 5.6 Conclusion -- Bibliography -- Part II Signal Processing for Sensor Networks -- 6 Detection in Sensor Networks -- 6.1 Centralized Detection -- 6.2 The Classical Decentralized Detection Framework -- 6.2.1 Asymptotic Regime -- 6.3 Decentralized Detection in Wireless Sensor Networks -- 6.3.1 Sensor Nodes -- 6.3.2 Network Architectures -- 6.3.3 Data Processing -- 6.4 Wireless Sensor Networks -- 6.4.1 Detection under Capacity Constraint -- 6.4.2 Wireless Channel Considerations -- 6.4.3 Correlated Observations -- 6.4.4 Attenuation and Fading -- 6.5 New Paradigms -- 6.5.1 Constructive Interference -- 6.5.2 Message Passing -- 6.5.3 Cross-Layer Considerations -- 6.5.4 Energy Savings via Censoring and Sleeping -- 6.6 Extensions and Generalizations -- 6.7 Conclusion -- Bibliography -- 7 Distributed Estimation under Bandwidth and Energy Constraints.

7.1 Distributed Quantization-Estimation -- 7.2 Maximum Likelihood Estimation -- 7.2.1 Known Noise pdf with Unknown Variance -- 7.3 Unknown Noise pdf -- 7.3.1 Lower Bound on the MSE -- 7.4 Estimation of Vector Parameters -- 7.4.1 Colored Gaussian Noise -- 7.5 Maximum a Posteriori Probability Estimation -- 7.5.1 Mean-Squared Error -- 7.6 Dimensionality Reduction for Distributed Estimation -- 7.6.1 Decoupled Distributed Estimation-Compression -- 7.6.2 Coupled Distributed Estimation-Compression -- 7.7 Distortion-Rate Analysis -- 7.7.1 Distortion-Rate for Centralized Estimation -- 7.7.2 Distortion-Rate for Distributed Estimation -- 7.7.3 D-R Upper Bound via Convex Optimization -- 7.8 Conclusion -- 7.9 Further Reading -- Bibliography -- 8 Distributed Learning in Wireless Sensor Networks -- 8.1 Introduction -- 8.2 Classical Learning -- 8.2.1 The Supervised Learning Model -- 8.2.2 Kernel Methods and the Principle of Empirical Risk Minimization -- 8.2.3 Other Learning Algorithms -- 8.3 Distributed Learning in Wireless Sensor Networks -- 8.3.1 A General Model for Distributed Learning -- 8.3.2 Related Work -- 8.4 Distributed Learning in WSNs with a Fusion Center -- 8.4.1 A Clustered Approach -- 8.4.2 Statistical Limits of Distributed Learning -- 8.5 Distributed Learning in Ad-hoc WSNs with In-network Processing -- 8.5.1 Message-passing Algorithms for Least-Squares Regression -- 8.5.2 Other Work -- 8.6 Conclusion -- Bibliography -- 9 Graphical Models and Fusion in Sensor Networks -- 9.1 Introduction -- 9.2 Graphical Models -- 9.2.1 Definitions and Properties -- 9.2.2 Sum-Product Algorithms -- 9.2.3 Max-Product Algorithms -- 9.2.4 Loopy Belief Propagation -- 9.2.5 Nonparametric Belief Propagation -- 9.3 From Sensor Network Fusion to Graphical Models -- 9.3.1 Self-Localization in Sensor Networks -- 9.3.2 Multi-Object Data Association in Sensor Networks.

9.4 Message Censoring, Approximation, and Impact on Fusion -- 9.4.1 Message Censoring -- 9.4.2 Trading Off Accuracy for Bits in Particle-Based Messaging -- 9.5 The Effects of Message Approximation -- 9.6 Optimizing the Use of Constrained Resources in Network Fusion -- 9.6.1 Resource Management for Object Tracking in Sensor Networks -- 9.6.2 Distributed Inference with Severe Communication Constraints -- 9.7 Conclusion -- Bibliography -- Part III Communications, Networking and Cross-Layered Designs -- 10 Randomized Cooperative Transmission in Large-Scale Sensor Networks -- 10.1 Introduction -- 10.2 Transmit Cooperation in Sensor Networks -- 10.2.1 Physical Layer Model for Cooperative Radios -- 10.2.2 Cooperative Schemes with Centralized Code Assignment -- 10.3 Randomized Distributed Cooperative Schemes -- 10.3.1 Randomized Code Construction and System Model -- 10.4 Performance of Randomized Cooperative Codes -- 10.4.1 Characterization of the Diversity Order -- 10.4.2 Simulations and Numerical Evaluations -- 10.5 Analysis of Cooperative Large-scale Networks Utilizing Randomized Cooperative Codes -- 10.5.1 Numerical Evaluations and Further Discussions -- 10.6 Conclusion -- 10.7 Appendix -- Bibliography -- 11 Application Dependent Shortest Path Routing in Ad-Hoc Sensor Networks -- 11.1 Introduction -- 11.1.1 Major Classifications -- 11.2 Fundamental SPR -- 11.2.1 Broadcast Routing -- 11.2.2 Static Shortest Path Routing -- 11.2.3 Adaptive Shortest Path Routing -- 11.2.4 Other Approaches -- 11.3 SPR for Mobile Wireless Networks -- 11.3.1 Broadcast Methods -- 11.3.2 Shortest Path Routing -- 11.3.3 Other Approaches -- 11.4 SPR for Ad-Hoc Sensor Networks -- 11.4.1 A Short Survey of Current Protocols -- 11.4.2 An Argument for Application Dependent Design -- 11.4.3 Application Dependent SPR: An Illustrative Example -- 11.5 Conclusion.

11.6 A Short Review of Basic Graph Theory -- 11.6.1 Undirected Graphs -- 11.6.2 Directed Graphs -- Bibliography -- 12 Data-Centric and Cooperative MAC Protocols for Sensor Networks -- 12.1 Introduction -- 12.2 Traditional Medium Access Control Protocols: Random Access and Deterministic Scheduling -- 12.2.1 Carrier Sense Multiple Access (CSMA) -- 12.2.2 Time-Division Multiple Access (TDMA) -- 12.3 Energy-Efficient MAC Protocols for Sensor Networks -- 12.4 Data-Centric MAC Protocols for Sensor Networks -- 12.4.1 Data Aggregation -- 12.4.2 Distributed Source Coding -- 12.4.3 Spatial Sampling of a Correlated Sensor Field -- 12.5 Cooperative MAC Protocol for Independent Sources -- 12.6 Cooperative MAC Protocol for Correlated Sensors -- 12.6.1 Data Retrieval from Correlated Sensors -- 12.6.2 Generalized Data-Centric Cooperative MAC -- 12.6.3 MAC for Distributed Detection and Estimation -- 12.7 Conclusion -- Bibliography -- 13 Game Theoretic Activation and Transmission Scheduling in Unattended Ground Sensor Networks: A Correlated Equilibrium Approach -- 13.1 Introduction -- 13.1.1 UGSN Sensor Activation and Transmission Scheduling Methodology -- 13.1.2 Fundamental Tools and Literature -- 13.2 Unattended Ground Sensor Network: Capabilities and Objectives -- 13.2.1 Practicalities: Sensor Network Model and Architecture -- 13.2.2 Energy-Efficient Sensor Activation and Transmission Control -- 13.3 Sensor Activation as the Correlated Equilibrium -- 13.3.1 From Nash to Correlated Equilibrium - An Overview -- 13.3.2 Adaptive Sensor Activation through Regret Tracking -- 13.3.3 Convergence Analysis of Regret-based Algorithms -- 13.4 Energy-Efficient Transmission Scheduling -- 13.4.1 Outline of Markov Decision Processes and Supermodularity -- 13.4.2 Optimal Channel-Aware Transmission Scheduling as a Markov Decision Process.

13.4.3 Optimality of Threshold Transmission Policies.