Edge Computing: A Primer (eBook)

Dr. Weisong Shi is a Charles H. Gershenson Distinguished Faculty Fellow and a Professor of Computer Science at Wayne State University. There he directs the Mobile and Internet Systems Laboratory, Connected and Autonomous Driving Laboratory,  Intel IoT Innovators Lab, and Wayne State Data Science Initiativeand Wireless Health Initiative, investigating performance, reliability, power- and energy-efficiency, trust and privacy issues of networked computer systems and applications. He founded the Metro Detroit  Workshop on Connected and Autonomous Driving (MetroCAD). Dr. Shi is one of the world leaders in the edge computing research community, and has been advocating Edge Computing (Fog Computing) since 2014,  a new computing paradigm in which the computing resources are placed at the edge of the Internet, in close proximity to mobile devices, sensors, end users and the emerging Internet of Everything.Jie Cao is Assistant Professor (Cybersecurity) in the School of Information Security & Applied Computing at Eastern Michigan University. Dr. Quan Zhang is a Data Engineer at Salesforce. He was previously a Graduate Research Assistant at Wayne State University.

Contents 6
1 Introduction 8
1.1 What Is Edge Computing 9
1.1.1 Why Do We Need Edge Computing 9
Push from Cloud Services 9
Pull from the Internet of Things 10
Change from a Data Consumer to Producer 10
1.1.2 Key Techniques that Enable Edge Computing 11
VMs and Containers 11
Software Defined Networking (SDN) 11
Content Delivery/Distribution Network (CDN) 11
Cloudlets and Micro Data Centers (MDC) 12
1.1.3 Edge Computing Definition 12
1.1.4 Edge Computing Benefits 12
1.1.5 Edge Computing Systems 13
1.2 Overview of the Book 14
References 15
2 EdgeOSH: A Home Operating System for Internet of Everything 17
2.1 Introduction 17
2.2 Related Work 18
2.3 EdgeOSH: Overview and Design 19
2.3.1 Overview 21
2.3.2 Design 22
2.4 Summary 23
References 24
3 Firework: Data Analytics in Hybrid Cloud-Edge Environment 25
3.1 Introduction 25
3.2 System Design 26
3.2.1 Terminologies 27
3.2.2 Architecture 28
3.2.3 Programmability 31
3.2.4 Execution Model Comparison 36
3.3 Implementation 36
3.4 Discussion 38
3.5 Summary 40
References 40
4 Distributed Collaborative Execution on the Edges and Its Application on AMBER Alert 42
4.1 Introduction 43
4.2 Motivation 44
4.2.1 AMBER Alert 44
Limitations of Edge Devices 45
Control of the Vehicle Tracking Area 46
4.2.2 Distributed Collaborative Execution on the Edge 47
4.3 AMBER Alert Assistant 47
4.3.1 Application Scenario 47
4.3.2 Application Design 48
Control Center 48
Task Receiver 49
Data Processor 49
4.3.3 Implementation Details 49
4.3.4 Task Scheduling 51
Distance-Related Diffusion 52
Location-Direction-Related Diffusion 53
4.4 Evaluation 54
4.4.1 Experimental Setup 54
4.4.2 Collaboration of Local Edge Nodes 55
4.4.3 Task Scheduling 58
4.5 Related Work 59
4.6 Summary 61
References 61
5 Challenges and Opportunities in Edge Computing 64
5.1 Programmability 64
5.2 Naming 65
5.3 Data Abstraction 66
5.4 Service Management 68
5.5 Privacy and Security 69
5.6 Application Distribution 70
5.7 Scheduling Strategies 71
5.8 Business Model 72
5.9 Optimization Metrics 72
5.10 Summary 74
References 75
6 Existing Edge Computing Tools 76
6.1 What Is Your Role in Edge Computing? 76
6.2 Virtualization 77
6.2.1 Virtual Machine and Container 78
6.2.2 Network Virtualization 81
6.3 Resource Management 83
6.3.1 Kubernetes and Docker 84
6.4 Developing Platforms for Edge Computing 86
6.4.1 Edge Analytics 86
6.4.2 Development Tools and Platforms 89
6.5 Summary 91
References 91
7 Conclusions 93