Logical Architecture of Edge Computing System: Cloud, Edge, and Device Collaboration, Definition and Relationship

introduction

With the advent of 5G and the Internet of Things era, the generation of massive data and task computing have had a great impact on existing networks. Although cloud computing based on the Internet provides extensive and on-demand access to virtual shared configurable computing and storage resources , is an excellent platform for processing massive data and computing tasks, but for high-speed access and ultra-low-latency applications such as online games, virtual reality, and ultra-high-definition video streaming in the 5G era and massive terminal interconnections, cloud computing cannot meet the requirements. its required.

At the same time, one of the key features of the next-generation Internet is that information is increasingly generated locally and consumed locally, with computing and storage resources available at a large number of edge devices. Therefore, in order to cope with the challenges of cloud computing, network pressure and improve user experience to meet business needs, the industry proposes to migrate the cloud computing platform to the edge of the network, that is, edge computing, and explore the inherent capabilities of the network to provide edge services near the data source to meet Key requirements in terms of agile connectivity, real-time optimization, intelligent applications, security and privacy.

1. "Cloud-edge-device" architecture

In the Internet of Things architecture for a new generation of information infrastructure, data processing and data-based intelligent services are becoming more and more important.
In the past two years, a relatively popular term called "edge computing" refers to placing simple processes that require real-time computing and analysis closer to terminal devices to ensure real-time data processing. Also reduces the risk of data transmission .
Recently, a new hot word has appeared called "cloud-edge collaboration". Its meaning is not very different from that of edge computing, but it emphasizes the architecture of "cloud-edge-device". The terminal is responsible for overall perception, and the edge is responsible for local data. Analysis and reasoning, while the cloud gathers all edge perception data, business data, and Internet data to complete industry and cross-industry situation awareness and analysis.

• "Cloud" is the central node of traditional cloud computing and the control terminal of edge computing;
• "Edge" is the edge side of cloud computing, which is divided into infrastructure edge and device edge;
• "End" refers to terminal equipment, such as mobile phones, smart home appliances, various sensors, cameras, etc.

AI-based intelligent services run through the entire "cloud-edge-end" architecture. At the sensing terminal, AI technology aims to improve the sensitivity and accuracy of overall perception, as well as the real-time nature of human-computer interaction and object-thing interaction. Simple logical reasoning is performed through the chip.

at the edge

AI technology is mainly responsible for collecting local data and related business data in the domain, completing the analysis and reasoning of the sensing data, and transmitting the relevant analysis results or models to the sensing terminal to achieve the collaboration between the sensing terminal and the edge cloud. At the same time, the edge The cloud and the edge cloud can also be shared through networking, sharing data, resources, algorithms, etc., to complete the mutual collaboration between the edge clouds.

In the clouds

Not only does it need to provide storage, computing, network, and security resources related to cloud computing similar to edge cloud, but it also needs to collect and fuse all data to provide intelligent services based on global data, including intelligent scheduling, operation and maintenance, and macro decision-making.

The cloud center is good at global, non-real-time, and long-term big data processing and analysis, and can play its advantages in long-term maintenance, business decision support and other fields.
Edge computing is more suitable for localized, real-time, short-period data processing and analysis, and can better support real-time intelligent decision-making and execution of local businesses. Edge computing and cloud center are in a complementary and synergistic relationship . Edge-cloud collaboration will amplify the application value of edge computing and cloud computing: edge computing is not only close to the execution unit, but also the collection and preliminary processing unit of high-value data required by the cloud, which can better On the contrary, the business rules or models optimized by cloud computing through big data analysis and processing can be delivered to the edge side, and edge computing runs based on new business rules or models.

It can be seen from the table that cloud computing and edge computing have their own advantages. The main advantages of cloud computing are massive computing and massive storage, high computing efficiency, and wide-area coverage. It is suitable for computing-intensive, non-real-time computing tasks and massive The parallel computing and storage of data can give full play to its advantages in the fields of long-term maintenance and business decision support, and the computing hardware is concentrated in the cloud computing center to implement centralized management, so there is no need to maintain computing hardware, data storage and related software locally .

The main advantage of edge computing is that widely distributed edge nodes provide real-time data processing. The process of edge computing is a user- and application-centric process, which makes up for the defects of delay and mobility in cloud computing, and is suitable for non-computing-intensive processing analysis and real-time intelligent decision-making of type, real-time, and mobility data, and as a new network paradigm, it can meet the unprecedented growth of computing demands in the 5G era and the continuous improvement of user experience quality. The cloud is also more secure.

However, the processing performance of the edge platform is usually not as good as that of the cloud platform, and it usually does not have enough memory and processors to process large amounts of data, so it cannot perform complex operations such as deep learning

In addition to cloud-edge collaboration, it can be seen from the figure above that the logical architecture focuses on the interaction and collaboration between the cloud, edge, and terminal parts of the edge computing system, including cloud-edge collaboration, edge-device collaboration, and cloud-edge-device collaboration. part.

• 1. Cloud and edge collaboration : realized through the control node of Kubernetes in the cloud part and the nodes run by KubeEdge in the edge part.

The Kubernetes control node follows the original data model of the cloud part, keeping the original control and data flow unchanged, that is, the node running on KubeEdge appears as a normal node on Kubernetes. Kubernetes can manage the nodes running on KubeEdge just like ordinary nodes.
The reason why KubeEdge can run on edge nodes with limited resources and uncontrollable network quality is because KubeEdge implements the Kubernetes cloud computing orchestration containerized application through CloudCore in the cloud part and EdgeCore in the edge part on the basis of the Kubernetes control node. sink.
As shown below:

• 2. Edge-device collaboration : jointly realized through KubeEdge, the edge part, and EdgeX Foundry, the end part.

As a management program running on edge nodes, KubeEdge is responsible for managing resources, running status, and failures of application loads on edge nodes. In some edge computing systems, KubeEdge provides the required computing resources for EdgeX Foundry services and is responsible for managing the entire life cycle of EdgeX Foundry services.
EdgeX Foundry is an IoT SaaS platform managed by KubeEdge. The platform manages various IoT end devices in the form of microservices. At the same time, EdgeX Foundry can collect, filter, store and mine data of various IoT terminal devices through the managed microservices, and can also issue instructions to various IoT terminal devices through the managed microservices to control the terminal devices. control.

• 3. Cloud, edge, and terminal collaboration : realized through the control node of the cloud solution Kubernetes, the edge solution KubeEdge, and the edge solution EdgeX Foundry.

References

• In-depth understanding of edge computing: cloud, edge, end working principle and source code analysis Written by Cui Guangzhang Zhijiang Laboratory engineer
• Internet of things data analysis system
• Explain the Logical Architecture of Edge Computing System: Cloud, Edge, and End Collaboration
• A Review of Cloud-Edge Collaborative Architecture Based on Software-Defined Networks, [Article ID: 1007-130X (2021) 02-0242-16] Li Bo, Hou Peng, Niu Li, Wu Hao, Ding Hongwei, Article ID: 1007-130X (2021) 02-0242-16