Wireless Sensor Networks (eBook)

Wireless Sensor Networks 1
Foreword 7
Preface 9
Acknowledgments 11
Contents 13
Contributors 15
Part I Wireless Sensor Networks Design for Deployment 17
1 Introduction 18
1.1 The Promise and the Challenge of Sensor Networks 18
1.1.1 Goals of this Book 20
1.1.2 How can this Book Help the Beginner WSN Practitioner? 20
1.2 Guide to Using this Book 22
1.2.1 Part I: Design Strategies for Deploying Sensor Networks 22
1.2.2 Part II: Case Studies 23
References 29
2 Learning from Deployment Experience 30
2.1 Illustrating the Problem: Three Deployments 30
2.1.1 Bangladesh Groundwater Monitoring, 20061 31
2.1.2 Peru Seismic Station Deployment, 20072 33
2.1.3 WaterWise: Monitoring an Urban Water Distribution System (2008)3 36
2.1.4 Discussion 38
2.2 WSN Design Strategies 39
2.2.1 From Smart Dust to Today's WSNs 39
2.2.2 Design Spaces and Design Views 42
2.2.3 Meeting Application Requirements 45
2.2.4 The Practical–Theoretical Divide: Open Research Questions and the Value of Deployment 47
2.3 Starting Points for Development: Existing Platforms 49
2.3.1 End to End WSN Solutions 49
2.3.2 Generic Solutions 51
2.3.3 Research Platforms 52
2.3.4 Discussion 54
2.4 Who Is Taking Off: the WSNs or the Market Analysts? 55
2.4.1 WSN Forecasts and Gartner's Hype Cycle 56
2.4.2 Current and Forthcoming WSN Research and Commercialization Opportunities 59
2.5 Summary of Strategic Recommendations 60
References 61
3 Designing for Deployment 66
3.1 Introduction 66
3.2 The Design for Deployment Process 67
3.3 Key Design Parameters 69
3.3.1 Sampling Rate and Data Rate 69
3.3.1.1 Dealing with a High Data Rate 70
3.3.2 Cost 71
3.3.3 Network Size and Density 72
3.3.4 Deployment Environment 73
3.3.5 Deployment Duration 74
3.3.6 Target Audience and Interaction Model 74
3.4 Iterative Deployment 75
3.4.1 The First Deployment Iteration 76
3.4.2 The Second Deployment Iteration 77
3.4.3 Subsequent Deployments 78
3.5 Lessons from the Field 79
3.5.1 Development Lessons 79
3.5.2 Deployment Lessons 81
3.5.3 Learning from the Deployment Experience 81
3.6 Summary 82
Reference 82
Part II Wireless Sensor Network Applications Case Studies 83
4 Volcano Monitoring: Addressing Data QualityThrough Iterative Deployment 84
4.1 Introduction 84
4.1.1 Overview of Seismoacoustic Monitoring 85
4.1.2 Opportunities for Wireless Sensor Networks 86
4.1.3 Overview of Three Deployments 87
4.1.4 Datum v Dataset Quality 89
4.1.5 Structure of this Chapter 90
4.2 Sensor Interface Board 90
4.2.1 2005 Board Redesign 92
4.2.2 Performance and Future Designs 93
4.3 Time Synchronization 94
4.3.1 Single-Hop Time Synchronization 95
4.3.2 Adaptation to Multi-Hop Using FTSP 95
4.3.3 Observed FTSP Instabilities 96
4.3.3.1 Timestamp Filtering 98
4.3.3.2 Timestamp Rectification 98
4.3.4 Evaluation 99
4.3.4.1 Lab Experiments 99
4.3.4.2 Comparison with Broadband Station 100
4.3.5 Lessons Learned 102
4.4 Event Detection 103
4.5 Addressing Storage and Bandwidth Limitations 105
4.5.1 Overview of Lance 106
4.5.2 Cardinal v Ordinal Utilities 107
4.5.3 Utility Functions 108
4.5.4 2007 Deployment 109
4.5.4.1 RSAM v EWMA Node Level Utility Calculator 109
4.6 Policy Modules 110
4.6.1 Example Policy Modules 111
4.6.2 Evaluation and Use at Tungurahua 112
4.7 Optimizing for Energy and Bandwidth Usage 114
4.7.1 Refocusing on Energy Usage 116
4.7.2 Cost Estimation 117
4.7.3 Lance Optimizer 118
4.7.4 Evaluation and Results 119
4.7.4.1 Simulation and Testbed Experiments 119
4.7.4.2 2007 Deployment Analysis 122
4.7.4.3 Results Summary 123
4.8 Conclusions and Future Work 124
References 125
5 VoxNet: Reducing Latency in High Data Rate Applications 127
5.1 Introduction 128
5.2 In-situ Acoustic Source Localization 129
5.2.1 Enabling In-Situ Automation and Interaction 130
5.2.2 Usage Scenario 130
5.2.3 Localization Algorithm and Components 131
5.3 The VoxNet Platform 132
5.3.1 System Architecture 133
5.3.2 Interaction Model 133
5.3.3 Hardware 134
5.3.4 Software 136
5.3.4.1 High Level Application Layer 136
5.3.4.2 Distribution and Interaction Layer 137
5.3.4.3 Platform Drivers and Services Layer 139
5.4 In-Situ Deployment 140
5.4.1 Discussion of Problems Encountered During Deployments 141
5.4.1.1 Deployment One 143
5.4.1.2 Deployment Two 144
5.4.1.3 Deployment Three 145
5.4.2 Deployment Four 146
5.4.3 Summary 147
5.5 Factors Affecting Timeliness 148
5.5.1 Event Frequency 149
5.5.2 False Detections 149
5.5.3 A Case for Dynamic Processing 151
5.6 Lazy Grouping 152
5.6.1 Lazy Grouping Experimentation 154
5.6.1.1 Experiment One: Data Reduction 154
5.6.1.2 Experiment Two: Latency Reduction 156
5.6.2 Discussion 157
5.7 Adaptation Policy 159
5.7.1 Evaluation 160
5.7.2 Gathering Empirical Data 161
5.7.3 Simulation 162
5.7.4 Performance 162
5.7.5 Accuracy 163
5.7.6 Effects of Varying Local Processing Time 165
5.7.7 Summary 167
5.8 Conclusion 167
References 168
6 Failure Is Inevitable: The Trade-off Between Missing Data and Maintenance 171
6.1 Introduction 171
6.2 Difficulties in Traditional Soil Monitoring 173
6.3 Wireless Sensor Network Deployment 176
6.3.1 Deployment 177
6.3.2 In-Situ Interaction 178
6.3.2.1 Mobility Usage Scenarios 179
6.3.2.2 Supporting In-Field Interaction 180
6.4 Problems in Data Analysis 181
6.5 Design of Vigilance 183
6.5.1 Estimating Missing Data 184
6.5.2 Application Uncertainty 186
6.5.3 Maintenance Suggestions 188
6.6 Implementation of Vigilance 189
6.6.1 Missing Data Estimation 189
6.6.2 Uncertainty Propagation 190
6.6.3 Maintenance Suggestions 190
6.7 Evaluation 190
6.7.1 Methodology 191
6.7.2 Prediction Accuracy 191
6.7.3 Coverage 194
6.7.4 Maintenance 196
6.7.5 Deployment Design Implications 198
6.7.6 System Performance 200
6.8 Conclusion 201
References 202
7 Cane Toad Monitoring: Data Reduction in a High Rate Application 205
7.1 Introduction 205
7.2 Background: Pilot Deployment 210
7.2.1 Goals and Challenges 210
7.2.2 Hardware 210
7.2.3 Frog Vocalization Recognition Algorithm 211
7.2.3.1 Spectrogram Generation 212
7.2.3.2 Machine Learning 213
7.2.3.3 Voting Process 213
7.2.3.4 Evaluation 213
7.2.4 Lessons 214
7.3 Iteration 2: Hybrid Sensor Networks 215
7.3.1 Goals and Accomplishments 215
7.3.2 Challenges 216
7.3.3 Wireless Sensor Hardware 216
7.3.4 Cane Toad Monitoring Prototypes 217
7.3.4.1 Pure: Stargates Only 217
7.3.4.2 Hybrid: Stargates and Mica2s 217
7.3.5 Evaluation 222
7.3.6 Lessons 224
7.4 Iteration 3: Lightweight Classification 224
7.4.1 Random Sampling 225
7.4.2 Envelope Extraction 227
7.4.3 Matched Filter 228
7.4.4 Evaluation 228
7.4.5 Lessons 229
7.5 Conclusions 231
References 232
8 ExScal: Dealing with Scale 235
8.1 The ExScal Application and Its Demands 236
8.1.1 ExScal Status and Some Project Facts 237
8.2 The Hardware Platform 238
8.3 Topology, Coverage, and Deployment 240
8.4 The Software Architecture 242
8.4.1 The Trusted Base and Its Deployment Components 242
8.4.2 The Location Component 243
8.4.3 Perimeter Security Component, OpAp 244
8.5 Management 250
8.6 Testing Prior To, and During the Final Deployment 251
8.7 Results and Conclusions 253
References 255
9 Glacier Monitoring: Deploying Custom Hardware in Harsh Environments 257
9.1 Motivation for the Glacsweb Project 257
9.2 System Design 259
9.2.1 Deployment 260
9.2.2 Probe Evolution 261
9.2.3 Base Station Evolution 264
9.2.4 WSN Advantages for the User 267
9.3 Summary of Recommendations and Observations 269
References 269
10 Adding the Human Element: Experience with a Wireless Patient Monitoring System 271
10.1 Introduction 271
10.1.1 Waiting Patients Need to Be Monitored 272
10.1.2 Brief Overview of SMART 273
10.2 Application Goals and Requirements 275
10.3 Component and Sub-System Selection 276
10.3.1 Patient Monitoring Node Selection 276
10.3.2 Caregiver Node Selection 277
10.3.3 Location Subsystem Selection 277
10.3.4 Central Server Selection 279
10.3.5 Building Through Integration 279
10.4 Software 280
10.5 Preparations for Deploying SMART in the Hospital 281
10.6 The Hospital Deployment Experience 283
10.6.1 Installation and Operation of the Location System 283
10.6.2 Battery Life and Power Management 284
10.6.3 Choosing a Location for SMART Central 284
10.7 Discussion 284
10.7.1 Hospital Deployment Life-Cycle 285
10.7.2 Preparing for the Reality of the Hospital Environment 285
10.7.3 Working with Stakeholders During the Deployment 286
10.8 Future Work: Disaster Management 287
10.9 Conclusions 287
References 289
Editors 
290 
Index 293