1st Edition
Networked Filtering and Fusion in Wireless Sensor Networks
By exploiting the synergies among available data, information fusion can reduce data traffic, filter noisy measurements, and make predictions and inferences about a monitored entity. Networked Filtering and Fusion in Wireless Sensor Networks introduces the subject of multi-sensor fusion as the method of choice for implementing distributed systems.
The book examines the state of the art in information fusion. It presents the known methods, algorithms, architectures, and models of information fusion and discusses their applicability in the context of wireless sensor networks (WSNs). Paying particular attention to the wide range of topics that have been covered in recent literature, the text presents the results of a number of typical case studies.
Complete with research supported elements and comprehensive references, this teaching-oriented volume uses standard scientific terminology, conventions, and notations throughout. It applies recently developed convex optimization theory and highly efficient algorithms in estimation fusion to open up discussion and provide researchers with an ideal starting point for further research on distributed estimation and fusion for WSNs.
The book supplies a cohesive overview of the key results of theory and applications of information-fusion-related problems in networked systems in a unified framework. Providing advanced mathematical treatment of fundamental problems with information fusion, it will help you broaden your understanding of prospective applications and how to address such problems in practice.
After reading the book, you will gain the understanding required to model parts of dynamic systems and use those models to develop distributed fusion control algorithms that are based on feedback control theory.
Introduction
Overview
Fundamental Terms
Some Limitations
Information Fusion in Wireless Sensor Network
Classifying Information Fusion
Classification based on relationship among the sources
Classification based on levels of abstraction
Classification based on input and output
Outline of the Book
Methodology
Chapter organization
Notes
Proposed Topics
Wireless Sensor Networks
Some Definitions
Common Characteristics
Required Mechanisms
Related Ingredients
Key issues
Types of sensor networks
Main advantages
Sensor Networks Applications
Military applications
Environmental applications
Health applications
Application trends
Hardware constraints
Routing Protocols
System architecture and design issues
Flooding and gossiping
Sensor protocols for information via negotiation
Directed diffusion
Geographic and energy-aware routing
Gradient-based routing
Constrained anisotropic diffusion routing
Active query forwarding
Low-energy adaptive clustering hierarchy
Power-efficient gathering
Adaptive threshold sensitive energy efficient network
Minimum energy communication network
Geographic adaptive fidelity
Sensor Selection Schemes
Sensor selection problem
Coverage schemes
Target tracking and localization schemes
Single mission assignment schemes
Multiple mission assignment schemes
Quality of Service Management
QoS requirements
Challenges
Wireless Sensor Network Security
Obstacles of sensor security
Security requirements
Notes
Proposed Topics
Distributed Sensor Fusion
Assessment of Distributed State Estimation
Introduction
Consensus-based distributed Kalman filter
Simulation example 1
Distributed Sensor Fusion
Introduction
Consensus problems in networked systems
Consensus filters
Simulation example 2
Simulation example 3
Some observations
Estimation for Adaptive Sensor Selection
Introduction
Distributed estimation in dynamic systems
Convergence properties
Sensor selection for target tracking
Selection of best active set
Global node selection
Spatial split
Computational complexity
Number of active sensors
Simulation results
Multi-Sensor Management
Primary purpose
Role in information fusion
Architecture classes
Hybrid and hierarchical architectures
Classification of related problems
Notes
Proposed Topics
Distributed Kalman Filtering
Introduction
Distributed Kalman Filtering Methods
Different methods
Pattern of applications
Diffusion-based filtering
Multi-sensor data fusion systems
Distributed particle filtering
Self-tuning based filtering
Information Flow
Micro-Kalman filters
Frequency-type consensus filters
Simulation example 1
Simulation example 2
Consensus Algorithms in Sensor Networked Systems
Basics of graph theory
Consensus algorithms
Simulation example 3
Simulation example 4
Application of Kalman Filter Estimation
Preliminaries
802.11 distributed coordination function
Estimating the Competing Stations
ARMA filter estimation
Extended Kalman filter estimation
Discrete state model
Extended Kalman filter
Selection of state noise statistics
Change detection
Performance evaluation
Notes
Proposed Topics
Expectation Maximization
General Considerations
Data-Fusion Fault Diagnostics Scheme
Modeling with sensor and actuator faults
Actuator faults
Sensor faults
The Expected maximization algorithm
Initial system estimation
Computing the input moments
Fault Isolation
System description
Fault model for rotational hydraulic drive
Fault scenarios
EM Algorithm Implementation
Leakage fault
Controller fault
Notes
Proposed Topics
Wireless Estimation Methods
Partitioned Kalman Filters
Introduction
Centralized Kalman filter
Parallel information filter
Decentralized information filter
Hierarchical Kalman filter
Distributed Kalman filter with weighted averaging
Distributed consensus Kalman filter
Distributed Kalman filter with bipartite fusion graphs
Simulation example A
Wireless Networked Control System
Sources of wireless communication errors
Structure of the WNCS
Networked control design
Simulation example B
Notes
Proposed Topics
Multi-Sensor Fault Estimation
Introduction
Model-based schemes
Model-free schemes
Probabilistic schemes
Problem Statement
Improved Multi-Sensor Data Fusion Technique
Unscented Kalman filter
Unscented transformation
Multi-sensor integration architectures
Centralized integration method
Decentralized integration method
Simulation Results
An interconnected-tank process model
Utility boiler
Notes
Proposed Topics
Multi-Sensor Data Fusion
Overview
Multi-sensor data fusion
Challenging problems
Multi–sensor data fusion approaches
Multi–sensor algorithms
Fault Monitoring
Introduction
Problem Formulation
Discrete time UKF
Unscented procedure
Parameter estimation
Improved MSDF techniques
Notes
Proposed Topics
Approximate Distributed Estimation
Introduction
Problem Formulation
Fusion with Complete Prior Information
Modified Kalman filter-I
Lower-bound KF-I
Upper-bound KF-I
Convergence
Fusion without Prior Information
Modified Kalman filter-II
Upper-bound KF-II
Fusion with Incomplete Prior Information
Modified Kalman filter-III
Approximating the Kalman filter
Lower-bound KF-III
Upper-bound KF-III
Fusion Algorithm
Evaluation and Testing
Simulation results
Time computation
Notes
Proposed Topics
Estimation via Information Matrix
Introduction
Problem Formulation
Covariance Intersection
Covariance Intersection Filter
Algorithm
Complete feedback case
Partial feedback case
Weighted Covariance
Algorithm
Complete feedback case
Partial feedback case
Kalman-Like Particle Filter
Algorithm
Complete feedback case
Partial feedback case
Measurement Fusion Algorithm
Equivalence of Two Measurement Fusion Methods
Tracking Level Cases
Illustrative example 1
Illustrative example 2
Testing and Evaluation
Fault model for utility boiler
Covariance intersection filter
Weighted covariance filter
Kalman-like particle filter
Mean square error comparison
Notes
Proposed Topics
Filtering in Sensor Networks
Distributed H∞ Filtering
Introduction
System analysis
Simulation example 1
Distributed Cooperative Filtering
Introduction
Problem formulation
Centralized estimation
Distributed estimation
Issues of implementation
Distributed Consensus Filtering
Introduction
Problem formulation
Filter design: fully-equipped controllers
Filter design: pinning controllers
Simulation example 2
Distributed Fusion Filtering
Introduction
Problem statement
Two-stage distributed estimation
Distributed fusion algorithm
Simulation example 3
Distributed Filtering over Finite Horizon
Introduction
Problem description
Performance analysis
Distributed H∞ consensus filters design
Simulation example 4
Notes
Proposed Topics
Appendix
A Glossary of Terminology and Notations
General Terms
Functional Differential Equations
Stability Notions
Practical stabilizability
Razumikhin stability
Delay Patterns
Lyapunov Stability Theorems
Lyapunov-Razumikhin theorem
Lyapunov-Krasovskii theorem
Some Lyapunov-Krasovskii functionals
Algebraic Graph Theory
Basic results
Laplacian spectrum of graphs
Properties of adjacency matrix
Minimum Mean Square Estimate
Gronwall–Bellman Inequalities
Basic Inequalities
Inequality 1
Inequality 2
Inequality 3
Inequality 4 (Schur Complements)
Inequality 5
Inequality 6
Bounding lemmas
Linear Matrix Inequalities
Basics
Some Standard Problems
S-Procedure
Some Formulas on Matrix Inverses
Inverse of Block Matrices
Matrix inversion lemma
Irreducible matrices
Biography
Magdi Sadek Mahmoud obtained BSc (Honors) in communication engineering, MSc in electronic engineering, and PhD in systems engineering, all from Cairo University in 1968, 1972, and 1974, respectively. He has been a professor of engineering since 1984. He is now a Distinguished University Professor at King Fahd University of Petroleum and Minerals (KFUPM), Saudi Arabia. He was on the faculty at different universities worldwide including Egypt (CU, AUC), Kuwait (KU), UAE (UAEU), UK (UMIST), USA (Pitt, Case Western), Singapore (Nanyang Technological) and Australia (Adelaide). He lectured in Venezuela (Caracas), Germany (Hanover), UK (Kent), USA (University of Texas at SA), Canada (Montreal, Alberta) and China (BIT, Yanshan). He is the principal author of thirty-four (34) books, inclusive book-chapters and the author/co-author of more than 510 peer-reviewed papers. He is the recipient of two national, one regional, and four university prizes for outstanding research in engineering and applied mathematics. He is a fellow of the IEE, a senior member of the IEEE, the CEI (UK), and a registered consultant engineer of information engineering and systems (Egypt). He is currently actively engaged in teaching and research in the development of modern methodologies of distributed control and filtering, networked-control systems, triggering mechanisms in dynamical systems, faulttolerant systems and information technology. He is a fellow of the IEE, a senior member of the IEEE, the CEI (UK), and a registered consultant engineer of information engineering and systems Egypt.
Yuanqing Xia was born in Anhui Province, China, in 1971 and graduated from the Department of Mathematics, Chuzhou University, Chuzhou, China, in 1991. He received his MS degree in Fundamental Mathematics from Anhui University, China, in 1998 and his PhD degree in Control Theory and Control Engineering from Beijing University of Aeronautics and Astronautics, Beijing, China, in 2001. From 1991-1995, he was with Tongcheng Middle-School as a teacher, Anhui, China. During January 2002 to November 2003, he was a postdoctoral research associate at the Institute of Systems Science, Academy of Mathematics and System Sciences, Chinese Academy of Sciences, Beijing, China, where he worked on navigation, guidance and control. From November 2003 to February 2004, he joined the National University of Singapore as a research fellow, where he worked on variable structure control. From February 2004 to February 2006, he was with the University of Glamorgan, Pontypridd, U.K., as a research fellow, where he worked on networked control systems. From February 2007 to June 2008, he was a guest professor with Innsbruck Medical University, Innsbruck, Austria, where he worked on biomedical signal processing. Since July 2004, he has been with the Department of Automatic Control, Beijing Institute of Technology, Beijing, first as an associate professor, and then, since 2008, as a professor. His current research interests are in the fields of networked control systems, robust control, sliding mode control, active disturbance rejection control and biomedical signal processing.
"By exploiting the synergies among available data, information fusion can reduce data traffic, filter noisy measurements, and make predictions and inferences about a monitored entity. Networked Filtering and Fusion in Wireless Sensor Networks introduces the subject of multisensor fusion as the method of choice for implementing distributed systems. The book examines the state of the art in information fusion. It presents the known methods, algorithms, architectures, and models of information fusion and discusses their applicability in the context of wireless sensor networks. ... After reading the book, readers will gain the understanding required to model parts of dynamic systems and use those models to develop distributed fusion control algorithms that are based on feedback control theory."
—IEEE Microwave Magazine, December 2015