边缘计算应用与其挑战
主讲人：唐宇涛
为什么研究边缘计算?
关于科研的思考
美国的研究者善于定义一个问题
比如大数据、元宇宙和边缘计算等
能够成为开创者, 成为”挖坑”的人
Pick up the “low hand fruits”
至少应该在一些子问题(subtask)上发掘
边缘计算产生的背景
Edge computing is a highly virtualized platform that provides compute, storage and network services between end devices and traditional cloud date centers.
边缘计算的网络结构
(Thousands) Cloud Big data processing & Data warehousing (far away)
(Millions) Edge Geographically close to end devices (close)
(Billions) Things Data sources & resource limited (close)
云计算的兴起 (2005 - 2015)
Signal: 2004
IBM 把其个人主机业务全部卖给了联想
IBM认识到高性能计算正在从个人转移到云上
- SaaS: Software as a Service (软件即服务)
- PaaS: Platform as a Service (平台即服务)
- IaaS: Infrastructure as a Service (基础设施即服务)
云计算的表现形式
- Data center is a computer (2008)
- Edge device
- Internet of Things (IoT)
- Mobile device
- Desktop/laptop
云计算与边缘计算
边缘计算就是云计算的延伸
云计算 & 大数据
“Big data is high-volume, high-velocity and high-variety information assets that demand cost-effective, innovative forms of information processing for enhanced insight and decision making.” — Gartner
Variety: Structured $\to$ Structed & Unstructured
Velocity: Batch $\to$ Streaming Data
Volume: Terabytes $\to$ Zettabytes
Memory duce
为什么需要边缘计算 (2015 - 2025)
来自云服务商的推力
- 低延时
- 高可用性
- 省带宽
来自终端设备的拉力
- 实时性需求
- 资源受限
- 安全/隐私的需求
边缘计算的代表性事件
Technology Preparation Period
- 1999, CDN
- 2005, Function Cache
- 2009, Cloudlet
- 2010, Mobile Edge Computing
- 2012, Fog Computing, Cloud-Sea Computing
- 2013, “Edge Computing” Concept
Rapid Growth Period
2015 Sep, ETSI, MEC white paper Nov, Open-Fog Consortium
2016
May, NSF, Edge Computing, Highlight Area
May, Edge Computing Definition
数据的分布
2022年
- 端侧15%
- 边缘35%
- 云中心50%
2030年
- 端侧10%
- 边缘70%
- 云中心20%
边缘计算在做什么?
Upstream (上行数据处理)
- 安全加密
- 数据预处理
- 本地训练
Downstream (下沉计算)
3层的边缘计算网络结构
边缘计算的主要玩家
Cloud Companies
Hardware Providers
Consortiums/Communities
CDN
Research
Communication Carriers
边缘计算的主要应用
边缘智能 (Edge Intelligence)
- Traffic Detection
- Autonomous Driving
- Location Tracking
- AR/VR
- Safety Monitor
边缘计算面临的挑战
最大难点: 异构性
总结
- 边缘计算的历史和背景
- 边缘时代的机遇