Components and Services for IoT Platforms (eBook)

Prof. Dr. Ing. habil. Michael Hübner is the Chair for Embedded Systems for Information Technology (ESIT) at the Ruhr University of Bochum (RUB) since April 2012. He received his diploma degree in electrical engineering and information technology in 2003 and his PhD de­gree in 2007 from the University of Karlsruhe (TH). Prof. Hübner did his habilitation in 2011 at the Karlsruhe Institute of Technology (KIT) in the do­main of reconfigurable computing systems. His research interests are in reconfigurable computing and particularly new technologies for adaptive FPGA run-time reconfiguration and on-chip network structures with application in automotive systems, incl. the integration into high-level design and programming environments.

Preface 6
Contents 8
Part I Platforms and Design Methodologies for IoT Hardware 11
1 Power-Shaping Configurable Microprocessors for IoT Devices 12
1.1 Introduction 12
1.2 Microprocessor Design Constraints for the Internet of Things 14
1.2.1 Microprocessors for the Internet of Things 15
1.3 The Role of Power Consumption in IoT Devices 17
1.3.1 Localized Effects of Power Consumption 18
1.3.2 Globalized Effects of Power Consumption 18
1.3.3 Reliability of Integrated Circuits 19
1.4 Low Power Design Opportunities for IoT Processors 20
1.4.1 Dynamic Voltage/Frequency Scaling 21
1.4.2 GALS Design Style and Its Advantages for IoT Processors 22
1.4.3 Body (Substrate) Biasing 24
1.4.3.1 Impact of Substrate Biasing on Latch-Up 25
1.4.3.2 Case Study 1: DVFS and Body Biasing on Standard “Bulk” Processes 26
1.5 CMOS FD-SOI Technology and Related Opportunities for IoT Processor Design 28
1.5.1 Application of Forward Body Biasing in FD-SOI 30
1.5.2 Case Study 2: FBB on FD-SOI Technology 31
1.6 Conclusions 33
References 33
2 Formal Design Flows for Embedded IoT Hardware 36
2.1 Introduction 36
2.2 Background and Existing Work 38
2.2.1 High-Level and Logic Synthesis 38
2.2.2 HLS Scheduling 39
2.2.3 Allocation and Binding Tasks 40
2.2.4 History of High-Level Synthesis Tools 41
2.2.5 Next Generation High-Level Synthesis Tools 43
2.3 Synthesis for Low Power 46
2.4 The C-Cubed Hardware Synthesis Flow 47
2.5 Back-End Compiler Inference Logic Rules 49
2.6 Inference Logic and Back-End Transformations 50
2.7 The PARCS Optimizer 53
2.8 Generated Hardware Architectures 53
2.9 Generated Hardware Execution Platform 56
2.10 Experimental Results and Conclusions 56
2.11 Conclusions and Future Work 60
References 60
3 AXIOM: A Flexible Platform for the Smart Home 65
3.1 Introduction 65
3.2 Smart Home Scenarios 67
3.3 The AXIOM Platform 69
3.4 Thread Management 71
3.5 The OmpSs Programming Model 72
3.5.1 OmpSs@FPGA 73
3.5.2 OmpSs@cluster 75
3.6 Some Initial Experiments 78
3.6.1 Methodology 78
3.6.2 Matrix Multiplication Benchmark 79
3.6.3 Experiments 79
3.7 Conclusions 80
References 81
Part II Simulation, Modeling and Programming Frameworks for IoT 83
4 Internet of Things Simulation Using OMNeT++ and Hardware in the Loop 84
4.1 Introduction 84
4.2 OMNeT++ 85
4.3 Related Work 86
4.4 Robots in Assisted Living Environments 87
4.5 Concept 88
4.5.1 Modified Scheduler 89
4.5.2 HiL Interfaces 89
4.5.3 Messages 90
4.6 Case Study 92
4.7 Conclusion 93
References 93
5 Towards Self-Adaptive IoT Applications: Requirements and Adaptivity Patterns for a Fall-Detection Ambient Assisting Living Application 95
5.1 Introduction 95
5.2 Requirements to System Design: The Fall-Detection Application 96
5.2.1 Case Study Overview 96
5.2.2 Requirements Gathering and Analysis 97
5.2.2.1 Overview of Modelling Languages 97
5.2.2.2 Requirements Engineering Methods 97
5.2.3 System Design 98
5.2.4 Adaptivity Requirements 99
5.3 Proposed Pattern-Based Approach: Current IoT 100
5.3.1 Patterns Applied 101
5.3.1.1 Adaptation Detection Pattern 101
5.3.1.2 Case-Based Reasoning Pattern 102
5.3.1.3 Centralised Architecture Pattern 103
5.3.2 System Design/Implementation Using the Proposed Pattern-Based Approach 103
5.4 Problems with Current Fall-Detection Systems and Pattern-Based Run-Time Adaptation: The Future IoT 105
5.5 Conclusions and Future Work 106
References 107
6 Small Footprint JavaScript Engine 109
6.1 Introduction 109
6.2 Analysis of Small Footprint Engines 111
6.2.1 Runtime Data Structure 112
6.2.2 Garbage Collection 113
6.2.3 ECMA-262 Coverage 113
6.2.4 Size and Performance 114
6.3 Proposed Engine 115
6.3.1 Overall Structure of Duktape 116
6.3.2 Improved String Concatenation and Join 116
6.3.3 Array Putprop Fastpath 117
6.3.4 Bitwise Operation/Compare with Native Stack 118
6.3.5 Indirect Threading 118
6.3.6 Lazy Built-in Objects Construction 119
6.4 Experimental Results 119
6.4.1 Environment 119
6.4.2 Performance 119
6.4.3 Footprint 120
6.4.4 Binary Size 120
6.5 Summary and Future Work 121
References 121
7 VirISA: Recruiting Virtualization and Reconfigurable Processor ISA for Malicious Code Injection Protection 123
7.1 Introduction 123
7.2 Threat Model 125
7.3 Motivation, Background, and Related Work 126
7.4 Proposed Code Injection Protection Architecture 128
7.4.1 VirISA Main Functionality 130
7.4.2 VirISA Architecture 131
7.4.3 VirISA Permutation Table Rules 132
7.5 Conclusions 134
References 135
Part III Opportunities, Challenges and Limits in WSN Deployment for IoT 137
8 Deployment Strategies of Wireless Sensor Networks for IoT: Challenges, Trends, and Solutions Based on Novel Tools and HW/SW Platforms 138
8.1 Introduction 138
8.2 Research Trends and State-of-the-Art Approaches 139
8.2.1 A New Perspective for On-Site WSN Deployments 140
8.3 General Overview of the Proposed System 141
8.4 On-Site Deployment and Maintainability Optimization Mechanism 143
8.4.1 Modeling of the Deployment and Maintenance Methodology 144
8.4.2 Deployment Optimization Mechanism 146
8.5 Wireless Mesh Networking Optimization Approach 148
8.6 System Implementation 151
8.7 Experimental Test Cases and Discussion 154
8.8 Conclusions and Future Perspectives 157
References 158
9 Wireless Sensor Networks for the Internet of Things: Barriers and Synergies 160
9.1 Introduction 160
9.2 WSN Programming Models and Tools 164
9.2.1 Low-Level Programming 164
9.2.1.1 Operating System-Level Programming 164
9.2.1.2 Virtual Machine or Middleware 165
9.2.2 High-Level Programming 165
9.2.2.1 Model-Based Development 166
9.2.2.2 Group-Level Programming 166
9.2.2.3 Network-Level Programming (Macroprogramming) 167
9.2.3 Evaluation of Existing WSN Programming Models and Tools 169
9.2.3.1 Low-Level Programming Evaluation 169
9.2.3.2 High-Level Programming Evaluation 170
9.3 WSN Hardware and Server-Side Support 170
9.3.1 WSN Hardware 171
9.3.1.1 Microcontroller 171
9.3.1.2 RF Device 171
9.3.1.3 RF Antenna 172
9.3.1.4 Energy Supply 172
9.3.1.5 Transducers 173
9.3.1.6 Package 173
9.3.1.7 Hardware Nodes 173
9.3.1.8 Server-Side Support 174
9.4 Semi-Automated WSN HW/SW Application Synthesis 174
9.4.1 Semi-Automated Development Flow Overview 175
9.4.2 Automated Hardware–Software Synthesis Tool Overview 176
9.4.3 Automated Synthesis Tool Input Interface 178
9.4.4 Structure of Top-Level and Library Components 180
9.4.5 System Synthesis Process 181
9.4.6 Synthesis Use for Legacy Designs 182
9.5 Synergies for WSN Development Tools and Platforms 185
9.6 Conclusion 187
References 188
10 Event Identification in Wireless Sensor Networks 192
10.1 Introduction 192
10.2 Wireless Sensor Network Characteristics 193
10.3 Event Detection Challenges in WSNs 195
10.4 Data Fusion Categorization 196
10.4.1 Data Fusion Algorithmic Approaches 197
10.4.2 Levels of Data Fusion 198
10.5 Classification of Event Detection Approaches 198
10.5.1 Model-Based Approaches 200
10.5.1.1 Arithmetic Model-Based Approaches 200
10.5.1.2 Map-Based Approaches 200
10.5.1.3 Probabilistic/Statistical Model-Based Approaches 201
10.5.2 Pattern Matching-Based Approaches 203
10.5.2.1 Prototype Matching Techniques 204
10.5.2.2 Signature Matching Techniques 204
10.5.3 Artificial Intelligence and Machine Learning-Based Approaches 205
10.5.3.1 Supervised Learning 205
10.5.3.2 Unsupervised Learning 208
10.5.3.3 Fixed Width Clustering 208
10.6 Performance and Behavioral Characteristics of Event Detection Techniques 209
10.6.1 Processing Model of Event Detection 210
10.6.2 Technique's Scalability 210
10.6.3 Sensor Data Types 210
10.6.4 Time Constrained Performance Demands 211
10.6.5 Density 211
10.6.6 Evaluation Approach 211
10.7 Conclusions 212
References 212
Part IV Efficient Data Management and Decision Making for IoT 216
11 Integrating IoT and Fog Computing for Healthcare Service Delivery 217
11.1 Introduction 217
11.2 Related Work 219
11.3 Fog Computing 222
11.3.1 Definition 222
11.3.2 Characteristics 223
11.3.3 Benefits of Integration with the IoT Technology for Healthcare Service Provisioning 224
11.4 Integration of IoT–Fog and Cloud 225
11.4.1 System Model 225
11.4.2 Fog Server Architecture 227
11.5 Use Case Scenarios 228
11.5.1 Daily Monitoring and Healthcare Service Provisioning 229
11.5.2 Extended eCall Service Delivery 231
11.6 Conclusion 233
References 234
12 Supporting Decision Making for Large-Scale IoTs: Trading Accuracy with Computational Complexity 237
12.1 Introduction 237
12.2 The Smart Thermostat Usecase 239
12.2.1 System Modeling 241
12.3 Challenges and Motivation 242
12.4 Support Decision Making with the Fuzzy Inference Systems 243
12.4.1 Fuzzy Inference System's Architecture 244
12.5 Communication Links 247
12.6 Evaluation 250
12.7 Conclusions 253
References 254
13 Fuzzy Inference Systems Design Approaches for WSNs 255
13.1 Data Classification in WSNs 255
13.1.1 Fuzzy Logic Critical Definitions 257
13.2 Design of a Fuzzy System 259
13.3 Literature Review of Applying Fuzzy Logic in WSN Scenarios 260
13.4 Designing a Health Status FIS for Wireless Sensor Networks Application Scenario 262
13.4.1 Challenges of Applying Data Mining Techniques in WSNs 262
13.4.2 Use Case Scenario 263
13.4.2.1 Healthcare Assessment Fuzzy Inference System 263
13.4.2.2 WSN QoS Assessment Fuzzy Inference System 265
13.4.2.3 Evaluation of Both Systems 268
13.4.3 Centralized Implementation of the HealthCare Assessment FIS in a WSN Platform 271
13.4.4 Decentralized Implementation of the HealthCare Assessment FIS in a WSN Platform 273
13.5 Evaluation Analysis of the Centralized and the Distributed Approach 274
13.5.1 Evaluation Setup 275
13.6 Conclusions 279
References 280
Part V Use Cases for IoT 282
14 IoT in Ambient Assistant Living Environments: A View from Europe 283
14.1 Introduction to AAL IoT Approach 283
14.2 AAL Involved Technologies 285
14.3 Research Efforts: Paving the Way of AAL IoT 286
14.4 Maturity Analysis of AAL IoT Developments and Platforms 293
14.5 Conclusions 298
References 298
15 Software Design and Optimization of ECG Signal Analysis and Diagnosis for Embedded IoT Devices 301
15.1 Introduction and Theoretical Background 301
15.1.1 The Electrocardiogram Signal 302
15.1.2 Arrhythmia Detection and MIT-BIH Database 303
15.1.3 Discrete Wavelet Transform 303
15.1.4 Support Vector Machines 304
15.2 Design and Exploration of ECG Analysis Flow 305
15.2.1 Overview of the Proposed ECG Analysis Flow 305
15.2.2 Building Blocks of the ECG Analysis Flow 306
15.2.2.1 Filtering 306
15.2.2.2 Heartbeat Detection 306
15.2.2.3 Heartbeat Segmentation 306
15.2.2.4 Feature Extraction 306
15.2.2.5 Classification 307
15.2.3 Design Space Exploration on SVM Classifier 307
15.2.3.1 Creating and Input Data Set for SVM Design Space Exploration 308
15.2.3.2 Design Space Exploration for SVM Tuning 310
15.3 Analysis Flow on Embedded IoT Platform 311
15.3.1 Employed IoT Platform 312
15.3.2 Mapping High Level Functionalities to Low Level Source Code 313
15.3.2.1 Filtering 314
15.3.2.2 Heartbeat Detection 314
15.3.2.3 Heartbeat Segmentation 314
15.3.2.4 Feature Extraction 314
15.3.2.5 Classification 315
15.3.3 Supporting Back-End Server Communication 316
15.3.4 Evaluation of ECG Analysis Flow on the IoT Node 317
References 323
16 Design for a System of Multimodal Interconnected ADL Recognition Services 325
16.1 Introduction 325
16.2 Recognizing Events in Audio Content 327
16.2.1 Feature Extraction 327
16.2.2 Feature Aggregation and Analysis 328
16.3 Recognizing Events in Visual Content 329
16.3.1 Preprocessing 330
16.3.2 Motion Detection 330
16.3.3 Silhouette Tracking 331
16.3.4 Behaviour and Activity Detection 332
16.3.5 Fused Audio-Visual Analysis 332
16.4 ADL Recognition Architecture 332
16.5 Conclusion 334
References 335
17 IoT Components for Secure Smart Building Environments 336
17.1 Introduction 336
17.2 Relative Technology Review 338
17.3 Architectural Considerations and Principal Components 343
17.4 Implementation Insights 346
17.5 Summary and Conclusions 350
References 351
18 Building Automation Systems in the World of Internet of Things 355
18.1 Introduction 355
18.2 Evolution of Building Automation Systems 356
18.3 Wired Protocols 357
18.3.1 BACnet 357
18.3.2 LON 357
18.3.3 KNX 358
18.4 Wireless Protocols 360
18.4.1 Z-Wave 360
18.4.2 ZIGBEE 360
18.4.3 Bluetooth Low Energy 360
18.4.4 EnOcean 361
18.4.5 Wi-Fi 362
18.4.6 THREAD 362
18.5 Requirements for Building Automation System 362
18.6 Building Automation Systems and “Internet of Things” 364
18.7 Building Automation Systems and the “KNX of Things” 366
18.8 KNX Web Service Gateway 369
18.9 KNX of Things in Practice: Design and Implementation of an Ambient Assisted Living Residence 371
18.10 Conclusion 374
References 375
Index 376