Core Network Evolution in 5G

2021-11-30 Author:By Wang Weibin, Guo Xuefeng Click:
Core Network Evolution in 5G - ztetechnologies
The browser version you are accessing is too low. To provide better experience, it is recommended that you upgrade the browser toEdgeBrowserOr, it is recommended.GoogleBrowser

Core Network Evolution in 5G

Release Date:2021-11-30  Author:By Wang Weibin, Guo Xuefeng  Click:

Introduction of 5G-Advanced
While 5G construction has entered the fast lane, the research on 6G network is also in full swing. The communications industry is tremendously impacting the development of various industries and fields, and has become a pioneer of social and economic growth. 
However, the development of the communications industry cannot be achieved overnight. Especially with the integration of multi-industries and multi-disciplines, the communication industry needs to develop collaboratively with more fields. 5G is undoubtedly a big leap of technology. Whether Massive MIMO, C-RAN, ultra-dense networking, millimeter wave, and in-band full duplex on the wireless side, or NFV/SDN, service-based architecture (SBA), network slicing and edge computing on the network side, are of revolutionary significance. Therefore, 5G R&D, deployment and application will face greater challenges. There is still a long way to go to the 5G vision. Some technologies also face new problems and limitations, and integration with industry partners needs to be deepened. 
It can be said that there is still a long way from 5G to 6G. Mobile communication is a process of iterative development. 5G needs further development and maturity, and 6G pre-research and reserve are also needed.
3GPP officially named the 5G evolution as 5G-Advanced (5GA) starting with R18 at the 46th Project Coordination Group (PCG) meeting in April 2021. 5G-Advanced oriented to 2025 defines new objectives and capabilities, and enables 5G to produce greater social and economic value through 5G evolution and enhancement. 
Core network, the management center of mobile network, plays a vital role in network access, security, connection, switching and routing. With the large-scale deployment of 5G networks, it is necessary to explore the development direction and evolution target of 5G-Advanced core network from two dimensions, i.e. 5G achievements and limitations, as well as 6G vision and key technologies, expand 5G achievements and application scenarios, and lay the foundation for 6G startup and long-term vision.

Insight into 5G-Advanced Core Network from 5G Status Quo
There are still many limitations and disadvantages in a 5G network.
The network architecture still needs to evolve to achieve a user-centric on-demand network. For example, the service-based architecture (SBA) is still limited to the control plane; there is a long way to dynamic elastic network; and the decentralized architecture needs to be studied in the data plane to reduce the risk.
Network capabilities still need to be enhanced to meet various needs of future industrial Internet. Network bandwidth and latency cannot meet the requirements of XR immersive experience and holographic communications. The limitations of uplink bandwidth and deterministic functions still restrict the development of industrial Internet applications, and mobility management cannot meet the need of high-speed service continuity.
—Network intelligence has a long way to go. Network data analytics function (NWDAF) and management data analytics function (MDAF) are still limited to data collection and analysis. The closed-loop artificial intelligence (AI) technology based on data perception, collection, analysis and decision-making needs to be further studied to achieve automatic and intelligent network o peration and maintenance. 
—The edge, pipe, cloud and terminal products need to be deeply integrated and collaboratively developed. In the future, ubiquitous network connections and service-based computing will become a reality. For users and applications, it is necessary to study the deep integration and collaborative development of computing and networks to maximize resource efficiency and optimize service experience.
—Security problems are becoming increasingly serious. The security of users and data has been paid more attention than ever before, and the trust and security of products and solutions have become basic requirements. With the applications of industrial Internet and the interconnection of more devices and heterogeneous networks, the security issues are more severe. 
The in-depth study of 5G-Advanced involves expanding application scenarios, optimizing existing 5G technologies, and enhancing new technologies and capabilities, as shown in Fig. 1.

Expanding Application Scenarios
5G-Advanced puts the expansion and integration of application scenarios in the first place. 5G is oriented to three major scenarios: eMBB, mMTC and URLLC. After more than two years of deployment in China, 5G eMBB applications have matured in the ToC market, but they are not satisfactory in the ToB market. When the originally promising industries such as the Internet of vehicles, industrial Internet, XR, smart city, and telemedicine are implemented, they are still faced the problems of protocol, networking, standard compatibility between private network, IoT terminals and communication network, data and network security, and network operation and maintenance. There are still challenges in network encapsulation, customization, manageability and controllability, as well as environmental adaptability, and the deep integration in more scenarios is far from enough. Under 5G-Advanced, more effort should be made to integrate with the industries, study and expand application scenarios, and truly achieve 5G enabling thousands of businesses.
Optimizing Existing 5G Technologies
5G-Advanced should be based on existing 5G technologies and promote their development and maturity for large-scale commercial use. Network slicing, non-public network (NPN), and 5G LAN are three powerful tools for the network enabling industry to provide private and customized networks, but they are not mature enough to be applied on a large scale. For example, network slicing is mainly implemented on the core network side, which is defined from the isolation policies of the control plane and user plane. With the development of cloud, software and programmable technology on the wireless access side, slicing can be extended to the wireless side to implement 5G end-to-end slicing. Moreover, there is still a lot of room for improvement in automated and intelligent slice operations. It is necessary to promote the standardization of CSMF, NSMF, NSSMF interfaces, improve NSSF, AMF, and TA-based RAN slice availability flows, and optimize slice SLA management. A large number of current 3GPP R18 proposals are also enhanced researches on existing 5G technologies, including the service-based UPF and N2 interface, SM/MM separation, and dual registration improvement in terms of network architecture, and NPN, 5G LAN, eNS and MEC capability enhancement in terms of industry enabling capability. New XR communications, multi-access coordination, and intelligence are also included.
Enhancing New Technologies and Capabilities
5G-Advanced should analyze its own capability weaknesses, and carry out research on new capabilities, new technologies and even new architectures. The high-reliability low-latency deterministic network is not only the basic requirement of industrial Internet, but also one of the necessary capabilities for 5G to enter the field of intelligent manufacturing. 5G needs to make breakthroughs in high-precision clock synchronization, deterministic gating control, latency, and TSN policy capability. With the development of computing networks, their combination with the core network needs to be studied to achieve integrated computing and network and efficient data forwarding and computing in mobile network. In addition, with the online operation of hierarchical network decoupling, service-based function, and network slicing, network intelligence is imperative. At present, TMF has defined the autopilot network and given the grading standard of network intelligence. 5GA also needs to continuously explore the application and implementation of intelligence in planning, construction, maintenance and optimization.

Insight into 5G-Advanced Core Network from 6G Vision 
IMT-2030 released the white paper on "6G Vision and Candidate Technologies" in 2021, proposing intelligent Internet of everything and digital twin as 6G vision. According to the white paper, the 6G network should have four major features: native AI, native security, multi-domain convergence, and deterministic network, including 12 candidate technologies in terms of architecture and capability. Among them, architecture technologies include distributed, integrated terrestrial and non-terrestrial network, native AI, native security, digital twin, and computing network. Capability technologies contain programmable network, integrated communications and sensing, deterministic network, data services, immersive sensing, and semantic communication. It can be seen that most of the technologies are not new, and have been applied in 5G. They should be enhanced in 5G-Advanced to lay a foundation for 6G (Fig. 2). There are also some technologies that have not been applied or are still in the early stage of research. A prospective research on 5G-Advanced should be made based on the combination of application scenarios and related technologies.

Integrated Communications and Sensing
Sensing is not a new concept in the telecom field, but its technology, scope, and application scenarios will continue to expand. Integrated communications and sensing must be a phased development. The ultimate vision of digital twin also depends on the sensing technology, i.e. the sensing and digitalization of the physical world. The core network and applications sense and cooperate with each other to provide efficient data transmission. The core network senses computing needs and resources to maximize resource efficiency. It cooperates with RAN to achieve high-precision terminal location and expand applications, and collaborates with AI to deliver intelligent network services such as gesture sensing, facial expression sensing, and environment sensing. It senses network topology, congestion and traffic in real time to monitor the network and predict traffic and fault. With the continuous integration and development of communications and sensing, digital twin virtual networks of NEs, subnets, and even the whole network can be achieved step by step. 
Customized Network 
The network customization capability is the basis of serving thousands of industries. However, the current customization capability based on network slicing cannot meet the industry requirements. In addition to the requirements for network indicators such as bandwidth, delay and jitter, industry applications also require customized protocols, terminal compatibility, service continuity, customized session type, network management and capability exposure. The customization capability of 5G-Advanced will be enhanced with the continuous development of the support layer, function layer and orchestration layer, flexibly providing differentiated network customization capability for industry applications. The capability includes the research on programmable chips and SRv6 and other programmable protocols in function and protocol customization, the further development of virtualization, cloud and software to provide basic capabilities for programmable networks, the evolution of service-based function, and the esearch on service/function chain orchestration and cloud network integration orchestration to promote sustainable development of programmability at the network level. 
Native AI and Security
AI and security technologies that have been introduced for a long time will complement and promote each other with telecom network technologies in 5G, 5G-Advanced or even in the future. The focus here is native, that is, the inherent, self-learning and self-growing capabilities corresponding to the external mode. Taking AI as an example, with the continuous growth of network data, only native and self-enhancement capabilities can meet the increasing challenges of data collection, analysis and security. In the implementation of specific native solutions, both native AI and native security are still in the preliminary research stage, and there is a long way to go. 

5G-Advanced is a cold reflection on actual 5G development. 5G still faces many difficulties and limitations, and its maturity, scenarios and value still have great room for improvement. 5G-Advanced is also a forward-looking research on 6G, conforming to the long-term evolution of 6G and cooperating with various industries to provide a technical reserve for 6G. 
As one of the two major fields of mobile communication, 5G-Advanced should be based on the expansion and core requirements of industry application scenarios, and make comprehensive efforts in standard research and evolution, technical capability improvement, actual deployment and implementation, so as to achieve the goal of enabling all industries, serving the society and promoting economic development.
ZTE has widely deployed its core network in 2G/3G/4G/5G networks. With its profound technology accumulation and R&D resources, ZTE has independently developed all 5G product families and is willing to explore the progress of 5G-Advanced with industry partners.