In 2019, edge computing attracted much attention from the industry, which once caused an investment boom in the capital market, and many people called 2019 the first year of edge computing. Rationally, there are inevitably some hype factors that contribute to such a hot situation, after all, the concept of edge computing has existed for many years. Of course, there is no doubt that the vigorous promotion of the industrial Internet and the continuous brewing of large-scale commercialization of 5G have made the entire industry full of confidence and expectations for the deep integration of IT and OT. In this case, perhaps edge computing is not hot or difficult.
In this issue of intelligent internal reference, we recommend the report from the Cloud Computing Open Source Industry Alliance, which analyzes the application needs and business models of cloud-side collaboration in typical scenarios with a rational and optimistic attitude, and paves the way for guiding the development of the industry and formulating relevant standards. If you want to bookmark the report of this article (cloud computing and edge computing synergy nine application scenarios), you can get it in the zhidong headline number reply keyword "nc383".
<h1 class="ql-align-justify" > a new wave of cloud-side collaboration</h1>
1. Edge computing is a new tentacle for cloud computing to expand to the edge side of distribution
The European Telecommunication Standardization Association considers edge computing to provide IT service environments and computing power at the edge of mobile networks, emphasizing proximity to mobile users to reduce latency in network operations and service delivery and improve user experience.
Gartner argues that edge computing describes a computing topology in which information processing, content capture, and distribution are placed closer to the source of the information.
Wikipedia sees edge computing as a way to optimize cloud computing systems, performing data processing at the edge of the network, close to the source of the data.
The Edge Computing Industry Alliance believes that edge computing is an open platform that integrates the core capabilities of network, computing, storage, and applications on the edge edge side close to the object or data source, and provides edge intelligence services nearby to meet the key needs of industry digitalization in terms of agile connectivity, real-time business, data optimization, application intelligence, security and privacy protection.
The Open Fog Computing Alliance believes that fog computing is a horizontal system-level architecture that brings computing, storage, control, and network functions closer to users in cloud-to-object continuity.
Although the various definitions of edge computing mentioned above are different in expression, they basically express a consensus: to provide services at the edge of the network closer to the terminal.
The "center-edge-end" form has been formed since the beginning of telecommunications. In the telecommunications era, program-controlled switching centers, program-controlled switches, and telephones have formed the original "center-edge-end" form; in the Internet era, data centers, CDNs, and mobile phones/PCs have continued this form;
In the era of cloud computing + Internet of Things, cloud computing centers, small data centers/gateways, and sensors have formed a new "cloud-edge-end" form.
▲ Schematic diagram of the development of "cloud-side-end"
If we seek the definition of edge computing only from the edge side itself, it seems difficult to fully accept this seemingly all-encompassing new thing. A relatively simple question is whether smart terminals, home gateways or other computing devices that have long existed around us have hidden the edge computing identity for many years.
In order to praise edge computing, some voices describe the concept of cloud computing as slightly narrow or deliberately put cloud computing as the opposite of edge computing. However, from the actual situation of technology or business evolution, edge computing is actually more of a new solution formed by the extension of cloud computing to the terminal and user sides. Edge computing itself is an extension of the concept of cloud computing, even if it is given an independent concept, it cannot be separated from cloud computing, and the two are dependent on each other and operate in synergy.
In this white paper, we believe that in scenarios such as the Internet of Things and large traffic, in order to meet the needs of wider connectivity, lower latency, and better control, cloud computing is moving towards a more global distributed node combination form, and edge computing is a new tentacle for its distribution to the edge side.
▲ Distributed cloud schematic
2. Typical products and business models of edge computing
The giant companies that dominate the cloud computing market rely on the first-mover advantage of cloud computing technology to sink cloud computing technology to the edge side, and take the opportunity to strengthen the edge side of artificial intelligence to vigorously develop edge computing. Relying on rich industrial scenarios, industrial enterprises carry out edge computing practices to strengthen field-level control. Telecom operators are embracing the 5G market opportunities and fully deploying edge nodes to lay a firm foundation for the layout of next-generation infrastructure.
ICT service providers are gradually expanding their cloud computing capabilities to edge devices. Internationally, cloud computing giants Amazon, Microsoft and Google have launched related edge computing products. Amazon launches AWS Greengrass feature software that extends AWS to devices to process data generated locally by terminals while still using shipping for management, data analysis, and persistent storage; Microsoft releases Azure IoT Edge side products that extend cloud analytics to edge devices to support offline use while focusing on AI applications at the edge; Google also launched the hardware chip Edge TPU and software stack Cloud IoT in 2018 Edge, which extends data processing and machine learning capabilities to edge devices, enabling devices to manipulate data from their sensors in real time and predict outcomes locally.
In China, Alibaba, Tencent, Baidu, Huawei, ZTE, Digital Dream Factory, Xinhuasan, etc. have also launched corresponding edge computing products.
Alibaba launches the Link IoT Edge platform. By deploying in smart devices of varying orders of magnitude and in end-side compute nodes. Through the definition model, it connects devices with different protocols and different data formats to provide secure, reliable, low-latency, low-cost, and easily scalable local computing services; Tencent launched CDN Edge for edge computing, sinking data center services to CDN edge nodes, corresponding end users with the lowest latency, while reducing the computing pressure and network load of user data centers.
Baidu launched the Intelligent Edge BIE, which expands cloud computing capabilities to user sites, provides temporary offline, low-latency computing services, and cooperates with the intelligent edge cloud management suite to form an end-cloud integrated solution of "cloud management, end computing".
In 2018, Huawei launched the IEF platform, which provides the ability to extend applications on the cloud to the edge by managing the user's edge nodes, linking data at the edge and in the cloud, and providing an edge computing solution with complete integrated services for enterprises to collaborate on edge and cloud.
ZTE launched an edge computing product, providing a full set of infrastructure from hardware to software, supporting a variety of edge computing system-level solutions, and providing a variety of resources for high-computing applications on the edge computing platform; DreamWorks launched a unified DT resource control product to provide "two fields + three forms" of integrated management and control capabilities, covering the central node field, the edge node field, and the unified control of the cloud platform, data middle office, and business middle office.
H3C launched UIS-Edge, a hyper-converged edge computing product that supports multiple hardware forms and deployment methods, extends cloud-native capabilities to the edge, provides perfect device access, edge computing, and cloud-side collaboration capabilities, and provides one-stop cloud-side integration solutions for enterprises.
Industrial enterprises rely on rich industrial scenarios to exert field-level application capabilities. Haier's one-stop device management platform COSMOEdge platform specially built for Internet of Things enterprises provides multi-source edge device access capabilities and powerful edge computing capabilities, supports the parsing of multiple industrial protocols, provides visual streaming pipelines, provides digital modeling and entity mapping, provides device-as-a-service application models, helps users quickly build Internet of Things applications, realizes digital production, and helps enterprises improve their efficiency; Root Cloud T-Box Vehicle IoT Box, Root Cloud Connector, Root Cloud IoT Agent Open Platform and a series of other covering mainstream industrial controllers and industrial protocol analysis, to achieve one-stop rapid access to all kinds of equipment in the industry, providing convenient, cheap and open device access solutions.
Telecom operators rely on 5G to fully deploy MECs. Mobile edge computing (MEC) is the use of wireless access networks to provide the services and cloud computing functions required by telecom users' IT nearby, and realize the flexible utilization of computing and storage resources. Multi-access edge computing (MAEC) extends edge computing from telecom cellular networks to other wireless access networks.
3. How cloud computing and edge computing work together
Take the IoT scenario as an example. Devices in the Internet of Things generate a large amount of data, data are uploaded to the cloud for processing, which will cause huge pressure on the cloud, in order to share the pressure of the central cloud node, the edge computing node can be responsible for the data calculation and storage work within their own scope. At the same time, most of the data is not a one-time data, those processed data still need to be gathered from the edge node to the central cloud, cloud computing to do big data analysis mining, data sharing, while the algorithm model training and upgrade, the upgraded algorithm pushed to the front end, so that the front end equipment update and upgrade, to complete the self-learning closed loop. At the same time, this data also needs to be backed up, and when there is an unexpected situation in the edge computing process, the data stored in the cloud will not be lost.
▲Schematic diagram of cloud-edge collaboration in the Internet of Things scenario
Cloud computing and edge computing need to work closely together to better meet the matching of various demand scenarios, so as to maximize the application value of cloud computing and edge computing. At the same time, starting from the characteristics of edge computing, real-time or faster data processing and analysis, saving network traffic, offline operation and support for resumable transmission, and higher security protection of local data are fully reflected in various scenarios of application cloud-side collaboration.
<h1 class="ql-align-justify" > two, cloud-side collaboration nine major application scenarios</h1>
This white paper will introduce the application of cloud-side collaboration in CDN, industrial Internet, energy, smart home, smart transportation, security monitoring, agricultural production and other scenarios, and analyze the prospects of cloud-side collaboration in healthcare, cloud games and other scenarios.
1. Application of cloud-side collaboration in CDN scenarios
With the current deployment of 5G, with AI technology, big data, cloud computing, IoT, etc., the internet of everything information age will let the Internet enter a new stage, the current stage of the CDN architecture has been unable to meet the application needs of the 5G era, CDN will usher in a new development of edge cloud + AI, in order to quickly respond to demand and achieve more transparency in service capabilities, service status and service quality. By deploying CDNs inside mobile networks, such as using edge cloud platforms to sink vCDNs (virtual Content Delivery Networks) into operators' edge data centers, the pressure on traditional networks will be greatly alleviated and the experience of mobile users' video services will be improved. Collaborative construction of CDNs based on cloud edges not only expands the pool of CDN resources on the basis of the central IDC, but also effectively uses the edge cloud to further improve the ability of CDN nodes to meet resource auto scaling.
▲VCDN implementation scenario based on edge cloud
CDN cloud-edge collaboration is suitable for localization + hot content frequent requests, and is suitable for supermarkets, residences, office buildings, campuses, etc. For recent hotspot videos and content, there may be frequent requests for localization, and after establishing a vCDN node locally through a remote content back-to-source, multiple requests for hotspot content in the local area can be distributed from the local node, which improves the hit rate, reduces the response latency, and improves the QoS metrics. Similarly, such processes can be applied to scenes such as 4K, 8K, AR/VR, 3D holograms, etc., and localization can quickly establish scenes and environments while improving the user experience, reducing vertigo and delaying stuttering.
2. Application of cloud-edge collaboration in industrial Internet scenarios
In recent years, with the successive introduction of relevant policy support by government departments and the continuous improvement of ecological construction, China's industrial Internet industry is developing rapidly. According to IDC, by 2020, more than 50% of the world's IoT data will be processed at the edge, and the industrial Internet, as an extension of the Internet of Things in the field of industrial manufacturing, also inherits the characteristics of massive heterogeneity of IoT data. In the industrial Internet scenario, the edge device can only process local data, can not form a global cognition, in practical applications still need to use the cloud computing platform to achieve information fusion, therefore, cloud-side collaboration is gradually becoming an important pillar to support the development of the industrial Internet.
Edge computing on the Industrial Internet works in tandem with cloud computing, with smart devices installed and connected in an edge computing environment capable of processing mission-critical data and responding in real time, rather than sending all the data over the network and waiting for a cloud response. The device itself is like a mini data center, with virtually zero latency since fundamental analysis is being performed on the device. With this new addition, data processing becomes decentralized and network traffic is greatly reduced. The cloud can collect this data later for a second round of evaluation, processing and in-depth analysis.
At the same time, in the field of industrial manufacturing, single point of failure is absolutely unacceptable in industrial-grade application scenarios, so in addition to the unified control of the cloud, the edge computing nodes in the industrial field must have certain computing power, be able to judge and solve problems independently, detect abnormal situations in time, better achieve predictive monitoring, and improve plant operation efficiency while preventing equipment failure problems. The processed data is uploaded to the cloud for storage, management, and situation awareness, and the cloud is also responsible for data transmission monitoring and edge device usage.
▲The industrial Internet uses edge cloud to achieve cloud-edge collaboration diagram
3. Application of cloud-edge collaboration in energy scenarios
Energy Internet is a new form of energy industry development in which the Internet is deeply integrated with energy production, transmission, storage, consumption and energy market, with the main characteristics of equipment intelligence, multi-energy collaboration, information accumulation, supply and demand dispersion, flat system, and open transactions.
In the process of upgrading the traditional energy industry to the energy Internet, the advantages of cloud computing and edge computing can be used to accelerate the upgrade process.
Taking the petroleum industry as an example, a large amount of production data is generated in all key links such as oil and gas exploitation, transportation, and storage. In the traditional mode, a large number of people are required to collect data regularly through manual meter reading, and monitor and check the equipment to prevent the occurrence of safety accidents. Meter reader regularly collects the data to report, and then by the data clerk to manually enter and analyze the data, one is that the labor cost is very high, the second is that the data analysis efficiency is low, the delay is large, and the status of the key equipment cannot be grasped in real time, and it is impossible to foresee safety events in advance to prevent accidents. The addition of edge computing nodes can realize real-time automated data collection and security monitoring of key equipment in key aspects of oil and gas exploitation through equipment such as temperature, humidity, pressure sensor chips and cameras with networking functions, and the raw data collected in real time is first collected to the edge computing node for preliminary calculation analysis, and the health status of specific equipment is monitored and related control is carried out. At this time, the data that needs to be interacted with the cloud is only high-value data after processing and analysis, which on the one hand greatly saves network bandwidth resources, and on the other hand, it also provides data pre-processing services for further big data analysis and data mining in the cloud, which avoids the multi-source heterogeneous data problem brought by a variety of collection equipment for the cloud.
▲The application of cloud-edge collaboration in the oil industry
In cloud-edge collaboration, terminal equipment or sensors are required to have certain computing capabilities, be able to process the collected data in real time, carry out local optimization control, automatic fault processing, load identification and modeling, interact with the cloud after processing and aggregation of high-value data, conduct network-wide security and risk analysis in the cloud, and carry out big data and artificial intelligence pattern recognition, energy saving and policy improvement. At the same time, if you encounter an area that the network does not cover, you can first process the data on the edge side, upload the data to the cloud in the case of a network, and store and analyze the data in the cloud.
4. Application of cloud-side collaboration in smart home scenarios
With the gradual development of information technology, the increasing perfection of network technology, the increasing abundance of applicable network carriers, and the gradual promotion of the strategy of large-bandwidth indoor network access, intelligent information services have become possible. The smart home comprehensively uses Internet technology, computer technology, remote sensing control technology, etc., to effectively combine family life such as family local area network, home device control, and family member information exchange to create a comfortable, convenient, safe and efficient modern home life.
In today's home intelligent information service into the home, how to simply access the smart home network of various heterogeneous household devices, and how users can conveniently use the functions in the smart home has become the focus of attention.
In the smart home scenario, the edge computing node (home gateway, intelligent terminal) has a variety of heterogeneous interfaces, including network cable, power line, coaxial cable, wireless, etc., and at the same time can also process a large number of heterogeneous data, and then upload the processed data to the cloud platform. Users can not only control home terminals by connecting edge computing nodes to the internet, but also access long-term data by accessing the cloud.
At the same time, the smart home cloud edge collaborates with the cloud service infrastructure based on virtualization technology, with diversified home terminals as the carrier, through the integration of existing business systems, the use of edge computing nodes will include household appliances, lighting control, multimedia terminals, computers and other home terminals to form a home LAN. The edge computing node is then connected to the WAN through the Internet (and through the 5G mobile network in the future 5G era), and then interacts with the cloud for data interaction, so as to realize functions such as electrical control, security protection, video surveillance, timing control, environmental detection, scene control, and video intercom.
In the future, cloud-side collaboration in smart home scenarios will be more and more valued by all parties in the industry chain, and telecom operators, home appliance manufacturers, and smart terminal manufacturers will explore in corresponding fields. In the near future, the family intelligent information service industry is not only limited to the control of household equipment, but also the industries of home energy, family medical care, home security, and family education will also be closely integrated with the application of family intelligence and become a member of the smart family.
▲Schematic diagram of the application of cloud-edge collaboration in smart home informatization
5. Application of cloud-edge collaboration in smart transportation scenarios
Vehicle-road coordination is an important development direction of intelligent transportation. Vehicle-road collaboration system is the use of advanced wireless communication and a new generation of Internet and other technologies, the comprehensive implementation of vehicles, vehicles, vehicles dynamic real-time information interaction, and on the basis of the collection and integration of dynamic traffic information in the whole space and space to carry out vehicle active safety control and road collaborative management, fully realize the effective coordination of people and vehicles, ensure traffic safety, improve traffic efficiency, thereby forming a safe, efficient and environmentally friendly road traffic system. According to the statistics of the Ministry of Public Security, by the end of 2018, the number of cars in China has exceeded 240 million, and the number of car drivers has reached 369 million. It is foreseeable that vehicle-road synergy has a huge market space in China, which provides a unique "testing ground" for the development and landing of intelligent transportation in China.
In the past, the focus of all parties on intelligent transportation was mainly focused on the car end, such as automatic driving, and the investment in research and development was mainly on the intelligence of the car, which put forward high requirements for the perception ability and computing power of the car, resulting in the high cost of smart cars. On the other hand, under the current technical conditions, the performance of autonomous vehicles in the traditional road environment is still unsatisfactory. Major manufacturers at home and abroad have gradually realized that roadside intelligence is indispensable for the realization of intelligent transportation, so in the past two years, they have invested in roadside intelligence, with the goal of achieving efficient interconnection and information sharing between people, vehicles and roads.
In practical applications, edge computing can be combined with cloud computing to integrate most of the computing load into the road edge layer, and use 5G, LTE-V and other communication means to interact with the vehicle in real time. In the future, the road edge node will also integrate local map system, traffic signal information, nearby moving target information and a variety of sensor interfaces to provide vehicles with a variety of services such as collaborative decision-making, accident early warning, and assisted driving. At the same time, the car itself will also become an edge computing node, working with the cloud edge to provide control and other value-added services for the vehicle.
The car will integrate lidar, camera and other sensing devices, and interact with the collected data with road edge nodes and surrounding vehicles, thereby expanding the perception ability and realizing the synergy between vehicles and vehicles, vehicles and roads. The cloud computing center is responsible for collecting data from a wide range of edge nodes, sensing the operation status of the traffic system, and issuing reasonable scheduling instructions for edge nodes, traffic signal systems and vehicles through big data and artificial intelligence algorithms, thereby improving the operational efficiency of the traffic system and minimizing road congestion.
▲Cloud-edge collaboration and vehicle-road collaboration reference framework
6. Application of cloud edge collaboration in security monitoring scenarios
At present, in the field of security monitoring, from the perspective of deployment and installation, the general traditional monitoring deployment adopts a wired method, the wired network covers all the cameras, the wiring cost is high, the efficiency is low, and it occupies a large number of wired resources. Using WiFi backhaul, WiFi stability is poor, the coverage is small, and a large number of routing nodes need to be supplemented to ensure coverage and stability. In the traditional way, surveillance video needs to be transmitted to the cloud or server through the bearer network and the core network for storage and processing, which not only increases the load of the network, but also makes it difficult to effectively guarantee the end-to-end latency of the service.
At the same time, a large number of camera acquisition terminals are equipped with strong data acquisition capabilities, on the one hand, the overall architecture of the camera puts forward higher requirements, how to ensure processing power and convenient installation in the case of fixed size and low power consumption, while on the other hand, as far as possible to ensure that the cost of camera collection end is low, is a more important issue.
Based on the above requirements, monitoring data can be diverted to edge computing nodes (edge computing service platforms), thereby effectively reducing network transmission pressure and end-to-end service latency. In addition, video surveillance can also be combined with artificial intelligence, equipped with AI artificial intelligence video analysis module on the edge computing node, for intelligent security, video surveillance, face recognition and other business scenarios, with low latency, large bandwidth, fast response and other characteristics to make up for the current AI-based video analysis generated by large latency, poor user experience problems, to achieve local analysis, rapid processing, real-time response. The cloud performs the training task of AI, and the edge computing node performs the inference of AI, and the two can realize local decision-making and real-time response, which can realize a variety of typical local AI applications such as expression recognition, behavior detection, trajectory tracking, hotspot management, and posture attribute recognition.
▲Schematic diagram of cloud edge collaborative application of intelligent security system
7. Application of cloud-edge collaboration in agricultural production scenarios
Smart agriculture is an advanced stage of agricultural production, which integrates emerging Internet, mobile Internet, cloud computing and Internet of Things technologies, relies on various sensing nodes and wireless communication networks deployed at the agricultural production site to achieve intelligent perception, intelligent early warning, intelligent decision-making, intelligent analysis, and expert online guidance for agricultural production, providing precision planting, visual management, and intelligent decision-making for agricultural production.
Take the smart greenhouse as an example: for the greenhouse with better conditions, electric roller blinds, exhaust fans, electric irrigation systems and other mechanical and electrical equipment are installed, and remote control functions can be realized through the cloud. Farmers can log in to the cloud system through mobile phones or computers to control the water valve, exhaust fan, roller switch in the greenhouse; you can also set the control logic in the cloud, the cloud will delegate the control logic to the edge control equipment, and the edge control device collects the air temperature, air humidity, carbon dioxide, light, soil moisture, soil temperature, outside the greenhouse temperature and wind speed through the sensing equipment in real time, and automatically opens or closes the roller shutter, water valve, fan and other greenhouse mechanical and electrical equipment according to the internal and external conditions.
▲Schematic diagram of cloud-edge collaboration application in smart greenhouses
8. Application analysis of cloud-side collaboration in cloud game scenarios
With the development of the Internet, and 5G networks have become a reality today, the term "cloud game" has begun to be used by more and more manufacturers, but also by more and more players are looking forward to. The so-called "cloud game" means that all games are running in the cloud server, and the cloud compresses the rendered game screen and transmits it to the user through the network to send it to the terminal. In the terminal, the user's gaming device does not require any high-end processor and graphics card, only need to have basic video decompression and instruction forwarding functions.
In 2018, telecom giants such as AT&T and Verizon, as well as IT giants such as Microsoft and Amazon, have announced cloud game-related tests or layouts. At the MWC in 2019, domestic mobile phone manufacturers OPPO and OnePlus also demonstrated related cloud gaming services. According to third-party forecasts, the global cloud gaming market will increase from $66 million in 2018 to $450 million in 2023, with a compound annual growth rate of 47%.
In the case of AR, the application needs to use the camera's view, positioning techniques, or a combination of both to determine where the user is and in which direction. After the location and orientation information is analyzed, the application can provide additional information to the user in real time. When the user moves, the information needs to be refreshed. Edge computing offloads computing tasks to edge servers or mobiles, reducing the average processing latency. The foreground interaction is placed on the cloud, and the background is handed over to the mobile terminal, ultimately enabling a complete AR experience.
▲Cloud edge collaboration applies block diagrams in cloud games
Zhi dong believes that there is a voice that edge computing is the opposite of cloud computing, which is actually wrong, and edge computing is more of a new solution formed by the extension of cloud computing to the terminal and user sides. Edge computing itself is an extension of the concept of cloud computing, even if it is given an independent concept, it cannot be separated from cloud computing, and the two are dependent on each other and operate in synergy. Cloud-side collaboration will become the mainstream model, and in this collaborative mode, cloud computing is advancing to a new form of more global distributed node combination.