“5G applications are going deep into all walks of life. “5G + edge computing + AI” is a new model for operators to help the digitalization and intelligence of vertical industries. This brings four new challenges to operators’ bearer networks. Operators need to focus on six key points to build a 5G MEC Ready bearer network.
5G applications are going deep into all walks of life. “5G + edge computing + AI” is a new model for operators to help the digitalization and intelligence of vertical industries. This brings four new challenges to operators’ bearer networks. Operators need to focus on six key points to build a 5G MEC Ready bearer network.
MEC is an important way for 5G to enable the digital transformation of thousands of industries
Application localization (data does not leave the campus), content distribution (large bandwidth), computing localization (low latency), promote the migration of service content/application/computing to the edge, driving MEC (Multi-access edge computing, Multi-access) EdgeComputing) and the 5G core network are moving down with the flow.
Figure 1 Business marginalization promotes the development of MEC and the downward shift of 5G core network
The 5G core network adopts a flexible architecture that separates the data forwarding plane (UPF: User Plane Function) and the control plane (SMF: Session Management Function), that is, UPF can be deployed flexibly on demand, and one control plane can manage many UPFs at the same time without affecting The performance of the 5G core network. This enables 5G to bring the following new advantages to MEC:
The core network UPF is moved down to the enterprise park, which can ensure that key business data does not leave the park, and it is easier to provide low-latency bearer solutions; operators can configure independent UPF for each user and customize wireless services for enterprise users;
Operators open 5G communication service programmability (such as positioning, wireless communication capabilities), which can be called by enterprise users and integrated into business systems, and enterprises can customize their own 5G innovative applications;
The sinking 5G MEC system and the enterprise network are directly interconnected, so that the applications distributed on the two network systems can be integrated and connected in real time, and many innovative applications can be made.
Therefore, 5G MEC is the key for 5G to better enable all industries at the edge of the network. It is a new model for operators to facilitate the digitalization and intelligence of vertical industries. It is also a touch point and key scene for operators to enter vertical industries.
Four major challenges for operators’ 5G MEC bearer network
The traditional 4G bearer network, because the traffic is mainly north-south, many operators adopt the L2+L3 method, which is no longer suitable for 5G MEC traffic localization requirements. 5G MEC poses four new challenges to operators’ bearer networks:
Figure 2 New challenges for operators’ 5G MEC bearer network
1) On-site MEC (deployed in the enterprise park) is a new application scenario. Operators need to have a low-latency connection between the base station and MEC in the enterprise park. Important business data of the enterprise cannot leave the park. This is for operators to access The net presents new challenges.
2) The downward shift of UPF in 5G MEC has led to the downward shift of 5G core network service ports (such as N4, N6, N9, 5GC OAM and other interfaces), making the original (wireless core network) L3 VPN on the backbone network in the 4G era move downward To the UPF access point; at the same time, a large number of distributed deployment of UPF increases the coverage of L3 VPN. Operators’ bearer networks need to support L3 VPN downshifting and extensive coverage of L3 VPN networks to meet the new challenges of massive deployment of 5G MEC.
3) MEC’s UPF needs to communicate with the 5G core network control plane and management system in the central cloud, and needs to meet the communication requirements of the telecom cloud; the application in the MEC may be a part of the cloud computing in the DC and requires interoperability and coordination. This poses a new challenge to the operator’s bearer network for edge-cloud collaboration.
4) MEC supports integrated access (fixed network and mobile) and provides seamless FMC services. The bearer network is required to provide MEC with a network connection across the mobile bearer network and fixed bearer network to complete the service intercommunication between the MEC and the central cloud and MEC , This poses new challenges in terms of network architecture and network interoperability for operators with dual planes of mobile bearer metropolitan area networks and fixed bearer metropolitan area networks. (Note: China’s three major operators have two metropolitan area network planes)
The four major challenges mentioned above are the new demands and new challenges brought about by 5G MEC. The data transmission model of the existing 3G/4G bearer network is no longer suitable for 5G services, and does not match the MEC localized service model. The bearer network should be an MEC Ready network that meets the above requirements.
Operators 5G MEC network architecture model and six key points of network construction
Figure 3 Carrier network architecture model from the MEC perspective
Due to the relatively large differences in the bearer network architecture of various operators, this article provides a bearer network architecture model from the MEC perspective:
The 5G core network has added many new service interfaces, and the traffic model has changed greatly: the red box in Figure 3 is the MEC system. The UPF in the MEC needs to pass through N3 (the data interface from the base station to the UPF) and N4 (the 5G core network control plane to the UPF control Interface), N6 (UPF to Internet data interface), N9 (UPF to Anchor UPF data interface), LI (monitor interface), OAM (UPF management interface) and related 5G system interconnection, VPN design and connection are complicated . The service interfaces in these 5G standards, except for N3, are all newly moved core network interfaces. The 5G bearer network architecture needs to focus on these new service interfaces and traffic models.
Edge-cloud collaboration: The MEP platform supports operations of operators or third-party services, which may be the real-time processing part (cloud edge) of complex cloud services, and requires real-time interaction with the central cloud. The business and communication systems in the upper part of Figure 3 are distributed from the metropolitan area network center to the regional center, and the deployment of each operator will be slightly different.
The aforementioned 5G MEC network communication model requires operators to focus on the following six key points when constructing MEC-ready bearer networks:
1) Shortest MEC access network: For N3 service flow from base station to MEC UPF, the operator should provide the shortest path service forwarding. In the scene of MEC, the N3 service flow should be directly forwarded to MEC through the mobile bearer router in the park. N3 service flow should not be allowed to detour in the operator’s network. There is no need for detours in the business flow. On the one hand, it is for low latency and saving operator network bandwidth, and on the other hand, it is to ensure that key business data of the enterprise does not leave the campus, as shown in Figure 4. This requires the MEC access point router to be able to forward data packets nearby. Therefore, the routing capability (L3 to the edge) of the operator’s access network equipment is a basic requirement for supporting the non-detour transmission of service flow.
Figure 4 MEC needs no detour, low latency access network
2) Low-latency fragmentation: In order to meet the low-latency and safety and reliability requirements of MEC applications, the carrier network needs to provide low-latency fragmentation network services for enterprise users. The MEC fragmented network includes wireless base stations, mobile bearer networks (between base stations and MEC), and UPF systems, that is, all network elements through which enterprise business flows to MEC. The fewer network elements that pass through, the lower the complexity of the fragmentation, which is more conducive to ensuring low transmission delay.
3) MEC’s external multipoint communication: The service flow between MEC and 5G core network (N4, OAM), MEP management platform, and other MECs are all multipoint-to-multipoint communication modes, and they all need to be supported by L3 VPN. The MEC bearer network requires the entire network to provide L3 VPN capabilities, including the access network, that is, the L3 VPN capabilities extend to the edge; and the L3 VPN needs to span multiple network segments such as metropolitan area networks and backbone networks. The MEC bearer network is much more complicated than the 4G bearer network in terms of the number of network elements (a large number of UPF moves down) or the network coverage (from access to backbone). Therefore, a flexible and powerful L3 VPN is required to support multi-point communication (as shown in the figure) 5).
Figure 5 The management and control service interface traverses multiple segments of the network
4) The integrated communication capability of routers in the MEC system: Small and micro MEC is the current mainstream mode of 5G MEC. Because of cost and communication requirements, MEC generally uses a one-layer integrated network model (as shown in Figure 6), unlike data centers. Complex multi-layer network architecture, but still need to provide all the required communication functions, such as the intercommunication between the devices in the MEC and the reliable connection of L2 and L3 between the VMs in the service, to complete the routing and intercommunication between the MEC and the external IP network (IP RAN) And reliable communication, and coordinated communication with the side cloud. A UPF NFV (such as UPF) can run multiple VMs to improve performance and reliability. MEC routers must provide multi-channel load balancing (ECMP). The current MEC requirement is 16-channel load balancing.
Figure 6 ECN model
5) Edge-cloud collaboration: MEC UPF is a sinking data plane of the 5G core network, and MEC applications are used as a sinking real-time processing unit of cloud services. Both require the carrier network to provide reliable cloud-side communication capabilities. Support edge-cloud collaboration in terms of automated deployment and operation and maintenance. The cloud-side collaboration of UPF can refer to the bearer solution of the telecom cloud.
6) Secure intercommunication between the two networks: Operators’ MEC networks need to communicate with their corporate networks so that companies can integrate 5G communication capabilities and MEC applications into their business systems. Nowadays, the routers in the MEC are generally used to communicate with the enterprise network. Network security is an issue of great concern to both the enterprise network and the operator’s network, and a firewall-based network security solution is required.
The Edge Computing Network Infrastructure Group (ECNI) helps the development of the MEC industry
In November 2016, the Edge Computing Alliance (ECC), the most influential industry alliance in the field of edge computing, was established. There are currently more than 250 corporate members, covering industrial manufacturing, smart cities, energy, ICT, and academic ecological partners.
In order to promote the development of the edge computing bearer network and the edge computing industry, ECC and Network 5.0 jointly established the Edge Computing Physical Infrastructure Joint Working Group (ECNI) to build consensus, combine the network 5.0 technology, and focus on the three major network integration scenarios of edge computing (operators) FMC scenario, operator network & industry park network convergence scenario, enterprise/industrial field network IT/OT convergence scenario), to build an edge computing network infrastructure technology system. ECNI uses technical specifications, white papers, Testbed/PoC, Best Practice, etc. as the carrier, MEC Ready test certification as the starting point, through the coordination of external organizations (CCSA, ETSI, ITU and other main positions) layout standards, promote the standardized construction of the bearer network to support Edge computing industry development. In November 2019, ECNI released a white paper on the operator’s edge computing network.
The 5G mobile communication system has made many improvements in supporting vertical industries, such as low-latency wireless communication, flexible core network architecture, super uplink, etc., which are the main features that distinguish it from 4G. MEC is a new model for operators to help digital and intelligent vertical industries. MEC is the beginning of the widespread distribution of intelligence on the network. In the future intelligent world of interconnected everything, intelligence based on edge computing will be dotted on the network. The 4G bearer network is constructed based on the 2C (for ordinary mobile phone users). The traffic model is a simple north-south, wireless core network centralized model. The MEC network requirements for vertical industries are not considered. Therefore, the construction of 5G MEC bearer network is not a simple 4G network. Bandwidth upgrade.