5G network introduces NFV, SDN, and SBA to enable three major scenarios: eMBB, mMTC and uRLLC. After R15, R16 and R17 enhance the support for vertical industry applications and network intelligence. The development of network and that of service are mutually reinforcing. The development of 5G network opens new horizons for industry applications. Applications like HD cloud gaming and industrial vision require 5G network to meet the requirements of low latency, deterministic performance, high edge computing power as well as high bandwidth. Therefore, 5G needs to continue its evolution. The next step in 5G evolution is 5G-Advanced (5GA) starting from R18, and the 5G core network will also continue its evolution.
Looking back, the wireless core network has been continuously integrating technologies from multiple fields. In 2G/3G network, CT is rapidly developing in wireless communications. Customized CT equipment supports the development of the voice service and value-added service in core network. With the development of IT technology, 4G core network introduces a full IP interface to support the IP broadband service and ICT integration. When entering the 5G era, the introduction of technologies such as VM, container and SBA further promotes the integration of ICT. With the development of big data and intelligence, network data analytics function (NWDAF) that provides network intelligence enters the CT field, marking the integration of DICT, which comprises of data technology (DT) and ICT. From R18 onwards, the 5G core network will further enhance operational technology (OT) as well as DT, IT and CT. The integration of OT, DT, IT and CT (ODICT) is based on the current 2G/3G/4G/5G integrated core network. Meanwhile, security is the cornerstone of the core network. ZTE puts forward the 5G-Advanced core network vision ODICT 4+1 (Fig. 1).

OT for Precise Network
The development of OT in 5G is mainly reflected in network determinism. A deterministic network provides bounded latency and jitter for end-to-end packet transmission, end-to-end ultra-reliable network transmission by using technologies such as multi-path transmission, and guaranteed deterministic transmission for deterministic flows in a mixed traffic network.
In the initial stage of 5G, technologies such as slicing, uRLLC and 5G LAN can be integrated to provide deterministic communications. R16/17 clearly enhances support for 5G-TSN integration, and defines a system architecture where the 5G system is perceived as a TSN bridge, which puts 5G into use for the industrial IoT. In the stage of 5G-Advanced, the 5G TSN technology will be further developed and enhanced in terms of reliability. 5G TSN supports the LAN-level deterministic communication with bounded latency. 5G-Advanced proposes the cross-domain deterministic communication where the core network will further connect with the transport network so that deterministic scenarios of end-to-end remote man-machine collaboration can be supported. 

DT for Intelligent Network 
DT is mainly reflected in the aspect of network intelligence. Network intelligence is the key development direction of 5G-Advanced and 6G.
In 5G-Advanced, network intelligence will play an important role in user experience optimization, efficient O&M and security guarantee. It can be used to assist in service QoS parameter adjustment, slice access control, user plane path selection and RAT/Frequency selection. It can also be used for intelligent analysis of network operations, which provides health score, anomaly detection and prediction and fault root cause analysis, and the corresponding capacity optimization, configuration optimization and resource scale-in/out. Intent-driven network is introduced to facilitate network planning, design and deployment, thus reducing the requirements for O&M personnel. The intelligent engine can intelligently analyze the mobility and interactive behaviors of the user terminals, identify user terminals with potential threats to protect network security.
NWDAF is a known intelligent network development direction. New technologies such as federated learning, intent-driven network and digital twin will also be applied in network intelligence. 

IT for Computing Power Network 
With the use of SDN, NFV, VM, container, and SBA architecture, IT has been deeply integrated with CT in current 5G core networks. In addition to the efficiency and agility of IT technology, the high performance, security, and reliability of CT are still guaranteed, which are the benefits of ICT integration.
Computing power network represents the further development of IT technologies in 5G network. In May 2021, China released the guiding opinions on accelerating the construction of a national collaborative innovation system for the integrated big data center, which first proposed promoting the use of computing resources as a service. The purpose of introducing computing power network technologies into 5G is to support national industrial upgrade, achieve digital transformation of the whole society, and achieve efficient and reasonable utilization of network and computing power resources through the use of intelligent technologies. At present, the three major operators in China have been intensively exploring computing power network to bring the value of network into full play. Many key technologies and solutions require in-depth analysis and evolution. The core network UPF is the first network hop. Technologies such as computing power routing, measurement and scheduling need to be integrated by the core network. Technologies such as network slicing, unified computing and network resource orchestration and intelligent O&M also need to evolve with the core network to maximize the utilization of network and computing power resources in the 5G-Advanced network.

CT for Ultra-Simple Network
The deployment of the 2G/3G/4G/5G core network is complicated because interconnection tests are required between different NEs and network functions. The 5G SBA architecture simplifies the communicating mode between network functions, but the interactions between network functions are not reduced. 5G-Advanced core network, on the premise of supporting the current deployment mode, needs to simplify deployment, simplify the design, and reduce signaling interactions between NEs. A unified architecture is used to support the coexistence of multiple access networks and interconnection among heterogeneous networks, to enable simplified access based on unified access protocols such as SRv6 and unified authentication mode. On the basis of current centralized management, the core network needs to be flatter (further integrated with the access network), more distributed, and adds autonomous domains (autonomous management within a certain region) to implement simplified core network deployment. At the same time, the SDK interfaces can be made available to third parties so that core network can be defined, which, when combined with the intelligent technology, enables simplified management. The ultra-simple core network is shown in Fig. 2. 

Trusted Network with Endogenous Security
The security of 5G-Advanced core network is equally important. Security and trustworthiness are the cornerstone of a network. A secure telecom network needs to provide trusted access, data and operation.
In 5G-Advanced core network, the identity authentication and trust management of network access equipment will be more complicated. The diversity of equipment IDs makes it difficult to adapt to the current authentication and authorization mode. It is necessary to study the feasibility of separating the equipment ID from the routing ID and the two-way authorization and authentication scheme based on the new ID system, so as to make the equipment ID unique, complete and tamper-proof, and to make the network access authentication trustworthy and secure. 
To enable trust management for network operation, the security monitoring process for the 5G-Advanced network should be visualized, and the network faults predictable, preventable and recoverable. The behaviors of network participants and network equipment operators have complete security elements with transparency and traceability, and can be predicted and controlled.
As for trust management for network data, when network data grows explosively with the rapid increase of data traffic, it is important to carry out data management, network operation indicator detection, and introduce network data security technologies like blockchain and federated learning on the premise of ensuring the data security of users and enterprises.