Release the white paper four challenges and six key points

330108-91-05
At the 2019 Edge Computing Industry Summit held on November 28, the working group issued the “Operator Edge Computing Network Technology White Paper”, which completed the first “small goal”. For the first time in the industry, this white paper summarizes the network requirements of various edge computing scenarios into seven areas. “Some of these needs can be done simply by changing the network, while others need to collaborate with network technology experts and academic institutions to promote the formation of a new network architecture to solve the problem.” Raber said.

These seven areas include: Multi-access, reliable connectivity, network edge cloud collaboration/cross-domain edge cloud collaboration on fixed or mobile networks, on-demand scheduling of computing power and the selection of optimal nodes for service processing requirements, interoperability and security trust requirements for the integration of carrier networks and campus networks, heterogeneous access networks on site, and deterministic delay/low delay, high bandwidth, and high concurrency network requirements.

330108-91-05
The white paper focuses on the construction of 5G MEC network for operators, puts forward four major challenges and six key points, and proposes to promote MEC Ready on the basis of 5G 5G Ready.

The first is the coverage challenge of L3VPN. The downshift of UPF requires the downshift of the service port of the wireless core network, which increases the coverage demand of L3VPN. The UPF needs to communicate with the 5GC control surface and management system in the central cloud, which puts forward the demand for edge cloud collaboration for the carrier network. The third is the challenge of fixed mobile convergence access. MEC needs to provide seamless FMC services, and ECA needs to provide multi-access and connect across wireless networks and fixed networks. The fourth is the on-site MEC challenge, the base station and MEC in the enterprise park need to be directly connected with low latency, and ensure that the important business data of the enterprise does not leave the park.

The six key points of building MEC Ready carrier networks are: shortest ECA, low latency fragmentation, integrated communication capabilities of routers in ECN, ECI multi-point communication, edge cloud collaboration, and secure interworking of carrier networks and enterprise networks.

Start from the architecture core sinking support customization

The white paper also proposed the Edge Computing Network technology system for the first time in the industry, dividing the relevant network infrastructure into three parts, namely ECA(Edge Computing Access, Edge Computing Access Network) and ECN(Edge Computing Network, ECN). Edge Computing Interconnect (ECI).

330108-91-05
The edge computing access network (ECA) is an important part of the edge computing network technology system that is different from the cloud computing network technology system. It needs to have the characteristics of integration, low delay, large bandwidth, large connection and high security, etc. Therefore, ECA needs to evolve and upgrade on the existing network, and its core idea is to shorten the distance of ECA. That is, the edge computing system is as close as possible to the user system both physically and logically to reduce traffic detour in the network.

Edge computing internal network (ECN) emphasizes the centralized management system with simplified architecture, complete functions, lossless performance and edge cloud collaboration. Because the scale of edge computing system is much smaller than that of cloud computing system, flat architecture and fusion architecture have become the development direction of ECN.

The Edge Computing Interconnection Network (ECI) is more complex than the data center Interconnection Network (DCI). Edge computing system involves connecting with various types of systems, including cloud computing system, other edge computing system, user-built system, etc. Therefore, ECI connects more objects and belongs to different operators, such as users themselves, cloud service operators, other edge computing operators, etc. At the same time, it is necessary to continue to provide low latency network performance on the basis of edge collaboration. This enables ECI to achieve a high degree of collaboration between edge, network and cloud from the level of infrastructure layout, management and control architecture and business products.

330108-91-05
“Before we looked at edge computing from a network perspective, now we look at the network from an edge computing business perspective, and we have redefined the ECA through that perspective.” Raber said.

Song Jun said that compared with the 4G bearer network, the core surface can be moved down to the park, which has several advantages: first, it can ensure that the key data does not leave the park, which is the requirement of most enterprise users; Second, after the core of the network is moved down, it is easier to realize the low-delay bearing scheme and the path is shorter; Third, operators can customize an independent server core network (UPF) for each enterprise, which can not be done before, equivalent to customized wireless services for each 5G customer. In addition, operators can open the programmable capabilities of 5G communication, such as positioning capabilities, users can call this information for programming, and where a robot goes in the park, you can know immediately.

After the edge computing is moved down, the operator network and the enterprise’s own network can be interconnected, and the entire application is pulled through the two networks. “Future applications will definitely be distributed in the enterprise network and edge computing network two parts, and their business is connected, the network is interoperable, so you can do a lot of innovative applications.” Song Jun said.