In the past two years, 5G and edge computing technology as a new type of infrastructure have received widespread attention from all walks of life, research, and use. What kind of super application scenarios will emerge and be the first to be applied on a large scale has become the hottest topic in the industry. " The scale of application must be strongly related to the technical characteristics of 5G and edge computing. " At the edge computing technology application innovation special session of the 2020 Yunqi Conference, Alibaba Cloud senior technical expert Yang Jingyu said.
Two-wheel drive edge computing for consumer and industrial video scenarios
Compared with 3G and 4G, the outstanding technical characteristics of 5G are large bandwidth, large connections, low latency and high reliability. The core network supports cloud deployment, and supports network slicing and edge offloading of terminal traffic. It is natively suitable for deployment of edge computing (MEC). . In the 5G era with such technical features, the upgrade of video experience will be activated first. Without the constraints of internet speed, the process of video consumption experience and industrial applications will usher in a leap-forward development.
On the consumer side, more high-definition, large-capacity, and interactive video media formats will be quickly applied. 4K/8K ultra-high-definition video, VR/AR immersive experience, 3A masterpiece ultra-real cloud games, etc. will enter the lives of the public; in the industry On the other hand, the application of video data in various industries such as transportation, education, new retail, industry, and urban governance is accelerating the large-scale implementation.
5G will be an era based on interactive video and high-definition video. In this era, the characteristics of large bandwidth and low latency of video pose new challenges to cloud computing infrastructure. How to make computing structured and distributed at the edge and center Designing application computing architecture based on experience and maximizing the reduction of core network transmission costs are issues that need to be considered and resolved in every scenario in the 5G era.
Yang Jingyu believes that as a 5G twin technology, edge computing has the characteristics of large number of nodes, wide distribution, closer distance to terminals, low access delay, and elastic scaling. It is the best solution for large-scale video applications and can be widely adapted. Meet the current computing and distribution needs of massive video data in various industries.
CDN on ENS: Alibaba Cloud edge computing along the way
Edge Node Service (ENS) is an edge computing product developed by Alibaba Cloud in 2017. Based on more than 2,800 CDN nodes around the world, Alibaba Cloud will gradually upgrade the technology of CDN node facilities and run on ENS through CDN. Alibaba Cloud edge facilities Gradually evolve from accelerated services to edge cloud-network integrated services. At present, Alibaba Cloud edge computing has achieved full coverage of 300+ nodes in provincial capital cities and third-tier cities in popular areas, and the calculation delay is controlled within 10 milliseconds.
At the 2019 Hangzhou Yunqi Conference a year ago, Alibaba Cloud once defined edge computing as urban computing. Its core concept is to gradually deploy ENS nodes in cities and connect with Alibaba public cloud infrastructure and ecological applications, focusing on providing local consumption. Application and industrial computing services. Today, one year later , a batch of PaaS platforms and applications such as Alibaba Cloud CDN system, RTC system, urban Internet of Things, industrial Internet platform, cloud games, cloud desktop, and Dharma Academy's visual computing platform have been smoothly deployed and run in ENS Above .
Yang Jingyu shared in his speech: At present, ENS already has multiple computing forms (virtual machines, containers, secure containers, digital twin computing), multiple storage services (highly available block storage, file storage, edge view caching, CDN caching) ), multiple network services (Internet IP, VPC network, Internet IP+VPC hybrid network), and integrated security (node security, network security, system security) and other comprehensive capabilities , inheriting many of the Feitian system with independent intellectual property rights of Alibaba Cloud Excellent technical characteristics.
In addition, ENS not only fully inherits the characteristics of the central cloud in terms of security and reliability, but also can produce and expand at any time as needed. At the application layer, the computing capabilities of nodes are scheduled based on vertical scenarios, the CDN is responsible for distribution, and the RTC system provides industry-leading audio and video capabilities. At the edge, a large number of live video on-demand PaaS platforms are deployed to minimize the cost of traffic. This system effectively collaborates with the central cloud and Feitian IoT operating systems to realize a complete cloud-side-to-end computing system and framework, providing Alibaba Cloud's ubiquitous computing capabilities, which can be widely used in various scenarios.
Focus on the comprehensive upgrade of cloud-network integration ENS infrastructure
As an extension of central cloud computing, edge computing naturally has the characteristics of decentralized deployment, wide area coverage, and complex environment. It is precisely because of these characteristics that edge computing and edge networks must form a cloud network to make computing more flexible and suitable for more Many application scenarios. At the meeting, Yang Jingyu introduced the two research directions of Alibaba Cloud edge computing cloud-network integration basic technology.
Collaborative network: Optimize data intercommunication between nodes, central cloud and terminal network
Based on a wide range of node capabilities, Alibaba Cloud has built an ENS collaboration network, which realizes the integrated collaboration of terminal-edge, edge-edge, and edge-center, shielding the complex network environment for upper-level business applications, and providing a gap between edge computing and central computing. Provide high-quality and secure network connection capabilities to improve data intercommunication efficiency.
For example, in many application scenarios such as online education and webcasting, there are requirements for internal data interaction between nodes and instance data intercommunication with the central cloud and terminal campus network. The collaborative network of ENS is based on three major aspects: edge networking, security acceleration, and network slicing. The technical features help the above scenarios perfectly meet the challenges. Among them, edge networking can open up instance private networks between edge nodes according to its own needs to form a software-definable VPC network; security acceleration has multi-path acceleration functions, and at the same time, the information security of the bearer network is realized through tunneling technology, effectively resisting nodes Network slicing is a software-defined private network service provided by ENS, which effectively guarantees communication quality. In the future, when the 5G network slicing service is opened, the ENS slicing network can be connected with the 5G slicing service. , And finally provide customers with end-to-end network slicing services.
Collaborative computing: Provide simple and easy-to-use edge cloud for terminal digital twin computing scenarios
As new technologies such as 5G and edge computing gradually solve technical problems such as transmission bandwidth and transmission delay, cloud computing of terminals has become a possibility. Yang Jingyu said: In the future, new terminals will be mainly used for information conversion from the physical world to the digital world. Traditional human-computer interaction functions such as keyboards and mice, displays, etc. belong to this category. Terminals will further develop towards 3D perception and immersive experience. . Each physical terminal creates a digital twin computing body in the digital world to complete the computing part of the terminal.
ENS collaborative computing is born for the terminal digital twin computing scenario. The cloud gaming cloud desktop PaaS platform only needs to register on the ENS to create a digital twin computing body, and pair it with the physical terminal’s identity ID to use the base station with the mobile phone. Similar experience, complete the nearby twin computing service, no need to care where the digital twin is deployed, all this is done by ENS.
ENS collaborative network and collaborative computing capabilities can solve the system coordination problem between cloud edge and edge nodes, freeing users from deploying and maintaining complex distributed computing facilities, and achieving true simplicity and ease of use.
Deeply cultivated in the edge computing of urban scene + urban application is the time
Relying on Alibaba's rich ecological application advantages, Alibaba Cloud continues to explore and pilot edge computing, and has already implemented good practices in multiple localized urban scenarios such as new transportation, media, education, home entertainment, and manufacturing.
In the field of smart transportation, Alibaba Cloud’s edge computing superimposed on AI capabilities has built a smart high-speed system for Gansu Public Transport Construction Group to realize high-speed video on-demand, live broadcast, intelligent detection, and data visualization, allowing traffic videos to be processed on the cloud nearby, and video delays are reduced At the level of 10 milliseconds, the AI calculation response time of traffic incident video is increased by 70%, and the efficiency of road safety incident processing is increased by 80%, which truly achieves global traffic situation awareness, makes the highway smarter, and greatly improves the service level.
In the field of Internet of Vehicles, Alibaba Cloud edge computing provides real-time video streaming, codec, audio return transcoding, live broadcast forwarding, image capture and other services nearby distribution and scheduling for Shouqi car-hailing, greatly reducing resource consumption and improving AI business efficiency Cost ratio, taking advantage of natural geographical reliability, can not only support the rapid implementation of vehicle-road collaboration and Internet of Vehicles, but also the best choice for V2X applications under 5G to open up decision-making, perception, and analysis links.
In the field of new retail, Alibaba Cloud Edge Computing and Hema have jointly created a new retail video big data system, which reduces the construction cost of Hemaar Cloud by more than 50%, improves store management efficiency, and allows consumers to enjoy a more convenient shopping environment. Facilitate the digital upgrade of the new retail industry.
In the field of education, Alibaba Cloud edge computing and the future will jointly build smart classrooms and smart classrooms. Tens of thousands of educational videos are connected to the cloud quickly. The system construction cycle is shortened to less than 2 days, and the digital application of the Internet in education scenarios is quickly upgraded. During the period, we will support'stop classes without suspension', and comprehensively improve the teaching coverage and classroom efficiency in the education field.
In the field of radio and television media, Wasu Media uses Alibaba Cloud Edge Node Service (ENS) as the platform base to create a national radio and television intranet CDN for flexible capacity expansion and convenient operation and maintenance. At the same time, it optimizes streaming based on ENS edge live broadcast transcoding capabilities Live media experience. In addition, relying on the edge computing power and edge storage capabilities of ENS, Wasu Media and Alibaba Cloud collaborated to form a smart radio and television city cloud, which will provide city-level cloud infrastructure for applications such as smart communities, smart parks, municipal management, and smart campuses.
In the field of IoT smart terminals, based on the terminal cloudification capabilities of Alibaba Cloud edge computing nodes, Tmall Wizard migrates terminal local applications to Alibaba Cloud edge cloud through millisecond-level real-time connections, achieving lower terminal hardware costs and more abundant Cloud application types and smoother user experience accelerate the creation of a pan-entertainment application ecosystem.
In addition, Alibaba Cloud has also expanded many 5G edge computing application innovations in vertical fields such as smart construction sites, bright kitchens, home entertainment, and cloud games. Yang Jingyu said: In the future, Alibaba Cloud will continue to use cities as the unit to accelerate the deployment of edge cloud infrastructure, create computing closer to users, empower the application scenarios of urban computing, and join hands with industry partners to welcome the arrival of the era of digital intelligence. .