Edge computing system design and practice: leading the new wave of technological innovation

---

With the rapid development of technologies such as the Internet of Things, big data, and artificial intelligence, the traditional centralized computing model can no longer meet the needs of modern society. Edge computing, as a new computing model, is gradually changing the way we live and work with its advantages of low latency, high bandwidth and high reliability. This article will conduct an in-depth discussion of edge computing systems from the aspects of edge computing concepts, design principles, key technologies, and practical applications.

1. The concept of edge computing

Edge computing is a new computing model that shifts computing tasks from the data center to the edge of the network so that data can be processed where it is generated, thereby greatly reducing data transmission delays and bandwidth consumption. Edge computing can not only improve the efficiency of data processing, but also protect user privacy and improve data security.

2. Design principles of edge computing

The design principles of edge computing mainly include the following points:

1. Distributed architecture: Edge computing systems usually adopt a distributed architecture to disperse computing tasks to various nodes in the network, thereby improving the processing power and reliability of the system.

2. Resource optimization: Edge computing systems need to optimize the configuration of network resources, including computing resources, storage resources, and communication resources, to meet the needs of different applications.

3. Security: Edge computing systems need to consider data security, including data encryption, access control and privacy protection.

4. Scalability: Edge computing systems need to have good scalability to adapt to changing application requirements and network environments.

3. Key Technologies of Edge Computing

The key technologies of edge computing mainly include the following aspects:

1. Edge node: The edge node is an important part of the edge computing system and is responsible for processing data in the network. Edge nodes can be physical devices or virtual devices.

2. Edge protocol: Edge protocol is a protocol used for data exchange and communication in edge computing systems. Common edge protocols include MQTT, CoAP and HTTP.

3. Edge services: Edge services are various services provided in edge computing systems, including data processing, data analysis, and data storage.

4. Edge applications: Edge applications are various applications running on edge computing systems, including smart homes, smart transportation, and smart manufacturing.

4. Practical Application of Edge Computing

Edge computing has been widely used in many fields, including the Internet of Things, industrial automation, intelligent transportation, and medical health.

1. Internet of Things: In the Internet of Things, edge computing can process data generated by devices in real time, reduce data transmission delays and bandwidth consumption, and improve the response speed and efficiency of devices.

2. Industrial automation: In industrial automation, edge computing can realize real-time monitoring and fault prediction of equipment, improving production efficiency and equipment service life.

3. Intelligent transportation: In intelligent transportation, edge computing can realize real-time positioning and path planning of vehicles, improving traffic efficiency and safety.

4. Healthcare: In healthcare, edge computing can achieve real-time monitoring of patients and disease prediction, improving the quality and efficiency of medical services.

In summary, edge computing, as a new computing model, is gradually changing the way we live and work with its advantages such as low latency, high bandwidth, and high reliability. However, edge computing also faces many challenges, including technical difficulties, security issues, and management issues. Therefore, we need continuous research and exploration to promote the development and application of edge computing.

---

"Edge Computing System Design and Practice"

Explore the cloud edge, reveal the principles, and practice edge technology: a guide that can truly guide engineers to implement edge computing projects, ready to use.

feature

(1) Comprehensive content: Basically covering all aspects of edge computing, readers can have a very comprehensive understanding of edge technology through this book.
(2) For all types of readers: It has in-depth theoretical and architectural research, as well as various practical methods, techniques and solutions, whether they are edge technology enthusiasts or professionals in this field Everyone can gain something from it.
(3) Close to reality: Many solutions and cases in the book come from projects and research problems that the author has personally experienced, and have certain inspiration and guidance for practical engineering applications.
(4) Practical and novel: All knowledge points and theories involved should be based on real applied technologies as much as possible, and at the same time, some of the latest research directions and results can be introduced in each part.
(5) Interesting: It not only introduces a lot of professional knowledge, but also intersperses a lot of interesting content, making this book both informative and interesting.

brief introduction

Currently, there are many edge computing-related books on the market that focus on theory, but this book particularly emphasizes the combination of theory and practice. Many cases, ideas and summaries in the book are derived from actual projects and practices. experience. This book not only explains what edge computing technology is (what), but also explains why (why) and guides how to do it (how).
This book provides a relatively comprehensive introduction and summary of the technical fields involved in edge computing. The book is divided into 10 chapters. Chapter 1 is an overall introduction; Chapters 2 to 5 mainly introduce the knowledge and technology at the infrastructure level involved in edge computing, including hardware, storage, communication and security; Chapters 6 to 9 It mainly introduces knowledge and technologies at the edge computing architecture and application level, including microservices, data processing, industrial Internet of Things, and machine learning; Chapter 10 introduces three typical edge computing open source frameworks.
This book is comprehensive in content, close to reality, practical and novel, and highly readable. It is especially suitable for reading and reference by engineers and researchers engaged in the field of Internet of Things and edge computing; it is also suitable for those who want to understand edge computing. Read for architects, engineers, and project managers; also suitable for computer and information technology students, as well as Internet of Things and edge computing technology enthusiasts.

About the Author

Yang Jian has more than ten years of experience in R&D, architecture and project management of large domestic and foreign technology companies. As a senior engineer, he participated in the design and development of enterprise-level information systems deployed and applied globally; responsible for the project management, design and implementation of Huawei's overseas supply chain information system, manufacturing information system and related Internet of Things and edge platforms; the last two In 2016, he participated in and was responsible for the research and development of several national key R&D projects. He has in-depth research and rich experience in edge computing, industrial Internet of Things and intelligent manufacturing.
Li Changle, Ph.D., senior engineer, has long been engaged in the application research of edge computing in new energy microgrids and energy storage systems, ship power system research, and products R&D, standards research and other work. It has undertaken and participated in more than 10 national, provincial and ministerial level scientific research projects, 1 industry group standard, and 10 authorized patents, including 8 invention patents. Selected into the 2021 "Shanghai Industrial Elite" high-level talents: young industrial talents.

Table of contents

第1章 边缘计算介绍
1.1　边缘计算简史 2
1.1.1　IT基础技术的演进历史 2
1.1.2　挺进边缘计算 4
1.2　云计算、IoT和边缘计算 7
1.2.1　近边缘端和远边缘端 8
1.2.2　边缘计算的应用场景 9
1.3　通信与硬件技术的发展对边缘计算的推动 11
1.3.1　计算单元和存储系统 13
1.3.2　能源管理和收集 15
1.3.3　通信技术 18
1.4　热门技术和边缘计算 20
1.4.1　5G技术和边缘计算 20
1.4.2　云计算、边缘计算和IoT 23
1.4.3　机器学习和边缘计算 24
1.4.4　移动边缘计算和移动云计算 26
1.5　云计算平台提供的边缘计算服务 26
1.5.1　AWS IoT Greengrass 27
1.5.2　阿里云Link Edge IoT 27
1.5.3　百度智能边缘 29
第2章 边缘计算的硬件
2.1　不同运算核心硬件在边缘计算中的应用 33
2.1.1　CPU与冯·诺依曼体系 33
2.1.2　GPU与并行处理 38
2.1.3　FPGA与ASIC 45
2.1.4　未来的新计算技术 49
2.2　边缘网关和边缘服务器 50
2.2.1　边缘网关 51
2.2.2　边缘服务器和边缘一体机 52
2.3　各种传感器技术 55
第3章 边缘计算存储系统设计和实现
3.1　边缘计算存储系统设计 61
3.1.1　边缘计算的分布式存储系统 61
3.1.2　分布式存储理论基础 62
3.2　开源分布式存储系统 66
3.2.1　直连式存储和集中式存储 66
3.2.2　大规模分布式存储技术 67
3.2.3　分布式存储系统总结 94
3.3　存储系统硬件技术的发展 94
3.3.1　早期存储硬件技术 94
3.3.2　固态硬盘（SSD）技术 95
3.3.3　未来的存储硬件 96
3.4　极端条件下的边缘数据存储 97
3.4.1　边缘计算和云存储能力的盲区 97
3.4.2　用卡车把数据送回去 98
第4章 边缘计算的通信
4.1　物联网和边缘计算的通信概述 101
4.1.1　对于边缘设备和物联网设备的通信要求 101
4.1.2　边缘计算底层通信协议的分类 102
4.1.3　应用层和消息层协议 104
4.1.4　通信相关标准组织介绍 105
4.2　边缘计算网络层通信协议介绍 107
4.2.1　RPL协议 108
4.2.2　LoRa协议 109
4.2.3　NB-IoT协议 110
4.2.4　LTE-M协议 112
4.2.5　Sigfox协议 113
4.3　现场边缘网络和通信 114
4.3.1　近距离网络通信协议之一：蓝牙技术 114
4.3.2　近距离网络通信协议之二：ZigBee 116
4.3.3　近距离网络通信协议之三：Wi-Fi 118
4.4　应用层协议 118
4.4.1　MQTT协议 119
4.4.2　CoAP协议 121
第5章 边缘计算的安全性
5.1　边缘计算面临的安全性挑战 125
5.1.1　边缘计算面临的重大安全挑战 125
5.1.2　信息安全领域是全新的战场 126
5.1.3　谈谈震网病毒 127
5.1.4　Mirai病毒 129
5.2　计算机安全的一些基本概念 131
5.2.1　计算机安全的本质 131
5.2.2　计算机系统安全的常用方法和概念 133
5.2.3　计算机加密算法介绍 136
5.2.4　网络安全技术 140
5.3　从可信计算到可信边缘计算 143
5.3.1　可信计算介绍 143
5.3.2　TPM 1.2、TPM 2.0和TPCM 144
5.3.3　基于TPM 2.0的可信计算 146
5.3.4　可信边缘计算 147
5.4　边缘计算安全问题分类 148
5.4.1　边缘接入安全问题 149
5.4.2　边缘服务器安全问题 150
5.4.3　物理安全问题 151
5.5　构建安全的边缘计算架构 152
5.5.1　边缘计算安全综合设计 153
5.5.2　边缘计算安全实践清单 154
第6章 边缘计算的微服务架构和消息机制
6.1　微服务架构介绍 157
6.1.1　典型的微服务架构 157
6.1.2　IoT+边缘计算的微服务架构 158
6.2　关于容器技术 159
6.2.1　容器技术（Docker）介绍 160
6.2.2　Docker引擎 160
6.2.3　虚拟机和容器的区别 162
6.2.4　进一步深入容器技术 164
6.3　微服务技术深度解析 165
6.3.1　软件开发模式和架构的回顾思考 165
6.3.2　微服务架构核心组件 168
6.3.3　P2P协议下的微服务通信 173
6.3.4　讨论Kubernetes和边缘计算 175
6.4　边缘计算的微服务架构设计 179
6.4.1　边缘计算微服务架构的考量 179
6.4.2　边缘计算架构设计 180
第7章 边缘计算的数据处理
7.1　边缘计算数据处理的价值 184
7.1.1　传统的数据分析流程 184
7.1.2　数据价值的思考 185
7.2　流数据采集和存储 186
7.2.1　流数据概述 186
7.2.1　设备接入和数据采集 188
7.2.3　边缘时序数据存储 192
7.3　时序数据处理 197
7.3.1　完整时序数据处理框架TICK 197
7.3.2　Prometheus和Grafana监控系统 201
7.3.3　流处理系统 204
7.4　时序数据分析和预测方法 207
7.4.1　时序数据的整理和可视化 207
7.4.2　时序数据的一些重要概念 211
7.4.3　统计时序预测方法 212
7.4.4　ARIMA模型训练和预测 215
第8章 工业边缘计算
8.1　工业边缘技术介绍 219
8.1.1　工业边缘计算的发展现状 219
8.1.2　工业边缘的应用场景 220
8.1.3　传统制造业信息系统改造 222
8.2　工业通信协议与接入技术 224
8.2.1　不同工业通信协议介绍 224
8.2.2　OPC UA协议及IT与OT的融合 229
8.2.3　工业通用接入技术 233
8.3　边缘计算基础设施和成本 236
8.3.1　边缘计算对基础设施的影响 236
8.3.2　边缘计算解决方案成本估算 239
第9章 机器学习和边缘计算
9.1　常用机器学习方法 242
9.1.1　机器学习的类型 242
9.1.2　机器学习的步骤和评估指标 244
9.1.3　基于概率的机器学习方法——朴素贝叶斯分类 247
9.1.4　数据简化和降维 250
9.1.5　决策树分类 254
9.1.6　传统的回归预测方法 257
9.2　深度学习方法介绍 262
9.2.1　多层感知机 262
9.2.2　CNN和RNN 264
9.3　强化学习 265
9.4　机器学习在边缘计算中的应用 274
9.4.1　工业边缘计算平台机器学习案例 274
9.4.2　强化学习在机器人控制中的应用 279
第10章 边缘计算开源框架
10.1　EdgeX Foundry 282
10.1.1　EdgeX Foundry简介 282
10.1.2　EdgeX Foundry的设备服务和核心服务 283
10.1.3　EdgeX Foundry的支持服务和应用服务 286
10.1.4　系统管理微服务 289
10.2　KubeEdge 290
10.2.1　KubeEdge简介 290
10.2.2　KubeEdge的安装和配置 292
10.2.3　KubeEdge对于K8s的改进 296
10.3　轻量级机器学习框架TensorFlow Lite 298
10.3.1　TensorFlow Lite的安装和运行 299
10.3.2　TensorFlow Lite模型的优化 301
10.3.3　给TensorFlow Lite模型添加元数据（Metadata） 304
10.4　边缘网络价值和未来的挑战 308
10.4.1　梅特卡夫定律和贝克斯特罗姆定律 308
10.4.2　未来信息技术发展的制约因素和边缘计算的关系 310

Foreword/Preface

Edge computing is the core technology of the Internet of Everything

Edge computing technology is developed on the basis of cloud computing and the Internet of Things, and is a natural evolution of the development of information technology infrastructure and architecture. In the development history of computer technology for decades, computer architecture has evolved from mainframe to client/server to cloud computing. Information technology has not only transformed all walks of life, but also created the emerging industry of the Internet, connecting people together. With the gradual popularization of Internet of Things technology and the rise of emerging technologies such as virtual reality, information systems have begun to gradually evolve from a cloud + terminal model to a combination of cloud and edge terminals. The edge computing technology will be the core of the future Internet of Everything.

Edge computing is not a single technology, but a comprehensive application system of a series of software and hardware technologies. Edge computing touches almost all aspects of information technology. Each area involved has a very wide range of professional knowledge, technology and experience. The value of edge computing technology is not the technology itself, but the synergistic effect produced by integrating edge computing into various industries. In the future, with the further development and maturity of IoT technology and edge computing technology, I believe that all industries will benefit from it; at the same time, more and more innovation opportunities will emerge in this field. Learning and understanding the knowledge and technology of edge computing should be a "compulsory course" for all IT people who keep up with the forefront of technology.

Edge computing is a comprehensive knowledge and technology system, and the business opportunities of edge computing are at the intersection of various industries and edge computing. In other words, what enables edge computing is not the technology itself, but the combination of this technology with other industries. We can see that the Internet of Things, edge computing and edge intelligence have blossomed in many industries, from smart manufacturing, smart transportation, smart agriculture, to smart medical care, autonomous driving, etc. Countless application scenarios and innovations are taking place, and it is expected that edge computing will create more invention opportunities in the future.

The development of various hardware, software and communication technologies is the technical foundation for the emergence and development of edge computing; at the same time, the development of technology also brings new needs and pain points, which becomes the driving force for the application and development of edge computing. Mobile Edge Computing (MEC) combined with communication technology will be an important support point for realizing 5G scenarios with low latency and highly controllable communication (uRLLC) in the future. MEC technology will definitely be needed to achieve 100ns latency wireless network in the future. . If cloud computing is the brain of modern information technology, then edge computing is the extension and tentacles of the brain, and many computing tasks will sink to the edge. Edge intelligence brings artificial intelligence into various areas of production and life, and will become the main technical means to combine and implement artificial intelligence technology with actual scenarios.

Target readers of this book

• IoT and edge computing technology enthusiast.
• Edge computing engineers and researchers.
• IoT engineers and researchers.
• Software developers and testers.
• Communications technology R&D personnel.
• Information technology practitioners.
• Computer and information science students from various institutions.

method of obtaining

To: http://product.dangdang.com/29646615.html
Kyoto: https://item.jd.com/14256744.html