Wireless Sensor Networks
An Information Processing Approach
- 1st Edition - July 6, 2004
- Latest edition
- Authors: Feng Zhao, Leonidas Guibas
- Language: English
Information processing in sensor networks is a rapidly emerging area of computer science and electrical engineering research. Because of advances in micro-sensors, wireless… Read more
Information processing in sensor networks is a rapidly emerging area of computer science and electrical engineering research. Because of advances in micro-sensors, wireless networking and embedded processing, ad hoc networks of sensor are becoming increasingly available for commercial, military, and homeland security applications. Examples include monitoring (e.g., traffic, habitat, security), industrail sensing and diagnostics (e.g., factory, appliances), infrastructures (i.e., power grid, water distribution, waste disposal) and battle awareness (e.g., multi-target tracking). This book introduces practitioners to the fundamental issues and technology constraints concerning various aspects of sensor networks such as information organization, querying, routing, and self-organization using concrete examples and does so by using concrete examples from current research and implementation efforts.
- Written for practitioners, researchers, and students and relevant to all application areas, including environmental monitoring, industrial sensing and diagnostics, automotive and transportation, security and surveillance, military and battlefield uses, and large-scale infrastructural maintenance
- Skillfully integrates the many disciplines at work in wireless sensor network design: signal processing and estimation, communication theory and protocols, distributed algorithms and databases, probabilistic reasoning, energy-aware computing, design methodologies, evaluation metrics, and more
- Demonstrates how querying, data routing, and network self-organization can support high-level information-processing tasks
Sensor networking and embedded systems professionals including development engineers, research scientists, system architects, etc., in a wide variety of companies from the defense industry to the home computing and electronics industry
1 Introduction 1.1 Unique Constraints and Challenges 1.2 Advantages of Sensor Networks 1.2.1 Energy advantage 1.2.2 Detection advantage 1.3 Sensor Network Applications 1.3.1 Habitat monitoring: wildlife conservation through autonomous, non-intrusive sensing 1.3.2 Tracking chemical plumes: ad hoc, just-in-time deployment mitigating disasters 1.3.3 Smart transportation: networked sensors making roads safer and less congested 1.4 Collaborative Processing 1.5 Key Definitions of Sensor Networks 1.6 The Rest of the Book 2 Canonical Problem: Localization and Tracking 2.1 A Tracking Scenario 2.2 Problem Formulation 2.2.1 Sensing model 2.2.2 Collaborative localization 2.2.3 Bayesian state estimation 2.3 Distributed Representation and Inference of States 2.3.1 Impact of choice of representation 2.3.2 Design desiderata in distributed tracking 2.4 Tracking Multiple Objects 2.4.1 State-space decomposition 2.4.2 Data association 2.5 Sensor Models 2.6 Performance Comparison and Metrics 2.7 Summary 2.8 Appendix A: Optimal Estimator Design 2.9 Appendix B: Particle Filter 3 Networking Sensors 3.1 Key Assumptions 3.2 Medium Access Control 3.2.1 The S-MAC Protocol 3.2.2 IEEE 802.15.4 Standard and ZigBee 3.3 General Issues 3.4 Geographic, Energy-Aware Routing 3.4.1 Unicast Geographic Routing 3.4.2 Routing on a Curve 3.4.3 Energy-Minimizing Broadcast 3.4.4 Energy-Aware Routing to a Region 3.5 Attribute-Based Routing 3.5.1 Directed Diffusion 3.5.2 Rumor Routing 3.5.3 Geographic Hash Tables 3.6 Summary 4 Infrastructure Establishment 4.1 Topology Control 4.2 Clustering 4.3 Time Synchronization 4.3.1 Clocks and Communication Delays 4.3.2 Interval Methods 4.3.3 Reference Broadcasts 4.4 Localization and Localization Services 4.4.1 Ranging Techniques 4.4.2 Range-Based Localization Algorithms 4.4.3 Other Localization Algorithms 4.4.4 Location Services 4.5 Summary 5 Sensor Tasking and Control 5.1 Task-Driven Sensing 5.2 Roles of Sensor Nodes and Utilities 5.3 Information-Based Sensor Tasking 5.3.1 Sensor selection 5.3.2 IDSQ: Information-driven sensor querying 5.3.3 Cluster leader based protocol 5.3.4 Sensor tasking in tracking relations 5.4 Joint Routing and Information Aggregation 5.4.1 Moving center of aggregation 5.4.2 Multi-step information-directed routing 5.4.3 Sensor group management 5.4.4 Case study: sensing global phenomena 5.5 Summary 5.6 Appendix A: Information Utility Measures 5.7 Appendix B: Sample Sensor Selection Criteria 6 Sensor Network Databases 6.1 Sensor Database Challenges 6.2 Querying The Physical Environment 6.3 Query Interfaces 6.3.1 Cougar sensor database and abstract data types 6.3.2 Probabilistic queries 6.4 High-level Database Organization 6.5 In-Network Aggregation 6.5.1 Query propagation and aggregation 6.5.2 TinyDB query processing 6.5.3 Query processing scheduling and optimization 6.6 Data-Centric Storage 6.7 Data Indices and Range Queries 6.7.1 One-dimensional indices 6.7.2 Multi-dimensional indices for orthogonal range searching 6.7.3 Non-orthogonal range searching 6.8 Distributed Hierarchical Aggregation 6.8.1 Multi-resolution summarization 6.8.2 Partitioning the summaries 6.8.3 Fractional cascading 6.8.4 Locality preserving hashing 6.9 Temporal Data6.9.1 Data aging 6.9.2 Indexing motion data 6.10 Summary 7 Sensor Network Platforms and Tools 7.1 Sensor Network Hardware 7.1.1 Berkeley motes 7.2 Sensor Network Programming Challenges 7.3 Node-Level Software Platforms 7.3.1 Operating system: TinyOS 7.3.2 Imperative language: nesC 7.3.3 Dataflow style language: TinyGALS 7.4 Node-Level Simulators 7.4.1 ns-2 and its sensor network extensions 7.4.2 TOSSIM 7.5 Programming Beyond Individual Nodes: State-centric programming 7.5.1 Collaboration groups 7.5.2 PIECES: A state-centric design framework 7.5.3 Multi-target tracking problem revisited 7.6 Summary 8 Applications and Future Directions 8.1 A Summary of the Book 8.2 Emerging Applications 8.3 Future Research Directions 8.3.1 Secure embedded systems 8.3.2 Programming models and embedded operating systems 8.3.3 Management of collaborative groups 8.3.4 Light-weight signal processing 8.3.5 Networks of high-data-rate sensors 8.3.6 Google for the physical world 8.3.7 Closing the loop with actuators 8.3.8 Distributed information architecture 8.4 Conclusion
"Wireless sensor and actuator nets, also known as motes and smart dust, are an emerging computer class based on a new platform, networking structure, and interface that enable novel, low cost, high volume, applications. This text and reference is a critical link to create this new class by covering the field of study for both practitioners and researchers. Unlike earlier computer classes that have been mostly evolutionary, motes require the “tall, thin man” that Carver Mead used to describe custom VLSI design. Motes system research and development require, deep knowledge of radio links, networks, operating systems, each application, and their interaction. Zhao and Guibas provide an excellent foundation for embarking on understanding and building these new systems."—Gordon Bell, Senior Researcher, Microsoft Corporation
"This book provides both an insightful overview of the emerging field of wireless sensor networks, and an in depth treatment of algorithmic signal and information processing issues. An excellent text for both professionals and students!"—Deborah Estrin, Center for Embedded Networked Sensing, UCLA
- Edition: 1
- Latest edition
- Published: July 6, 2004
- Language: English
FZ
Feng Zhao
Feng Zhao is a senior researcher at Microsoft, where he manages the Networked Embedded Computing Group. He received his Ph.D. in Electrical Engineering and Computer Science from MIT and has taught at at Stanford University and Ohio State University. Dr. Zhao was a principal scientist at Xerox PARC and directed PARC’s sensor network research effort. He is serving as the Editor-In-Chief of ACM Transactions on Sensor Networks.
Affiliations and expertise
Senior Researcher, Microsoft Research, Redmond, WA, USALG
Leonidas Guibas
Professor Guibas heads the Geometric Computation group in the Computer Science Department of Stanford University, where he works on algorithms for sensing, modeling, reasoning about, rendering, and acting on the physical world. He is well-known for his work in computational geometry, computer graphics, and discrete algorithms. Professor Guibas obtained his Ph.D. from Stanford, has worked at PARC, MIT, and DEC/SRC, and was recently elected an ACM Fellow.
Affiliations and expertise
Geometric Computing Group, Stanford University, Stanford, CA, USARead Wireless Sensor Networks on ScienceDirect