Intelligent Response to Changes of 5G Cloud Core Network Delivery and Operations

In the evolution to 5G core network, the software and hardware architecture of telecom equipment has undergone tremendous changes from traditional to virtual, cloud-based devices. The telecom industry has introduced not only IT software architectures but also new ideas and methods. While bringing a lot of convenience to telecom applications, the decoupling of three layers (cloud platform, core network services and hardware infrastructure) based on the OpenStack open-source technology has also made a great impact on network delivery and operations, and has therefore promoted their intelligent changes. 

Automatic One-Click Deployment to Solve Difficulties in Network Delivery  
Coexisting with 2G, 3G and 4G networks, 5G networks are commercially deployed on a large scale. To reduce deployment difficulties and improve delivery efficiency under the existing manpower and network conditions, an end-to-end automatic core network delivery solution has become an urgent need. ZTE has developed its NFV auto integration center (AIC) that can fully meet the needs of operators. 
By integrating and solidifying the NFV integration process, AIC automates the design and planning, deployment, and acceptance tests of 5G virtual core networks, significantly shortens the network construction time, and achieves "zero distance" between tool functions and customer needs in the whole life cycle of network integration through a friendly user interface. 
The AIC platform provides end-to-end one-stop tool support for virtual core network construction involving network planning and design, scheme verification, network deployment and integration tests, and realizes automatic integration of various scenarios such as NFVI, MANO and VNF. With the AIC tools, operators can automate the planning, deployment and test of their virtual networks, which greatly improves the efficiency and quality of NFV integration. The AIC platform can also considerably reduce costs and difficulty in the integration. This helps operators accelerate their NFV construction and achieve rapid network transformation with simplicity.

Intelligent and Rapid Fault Location to Solve Difficulties in Network Operations
To change the inherent operations mode of traditional equipment including restart, switch-over, and board removal and re-installation, there is an urgent need of automatic and intelligent technical support for network operations. This can make full use of virtual features to liberate people from heavy operations work, rapidly locate problems in the system, and solve them. After the network is delivered and put into formal operation, how can operators rapidly demarcate and locate the problems or failures that will occur? ZTE has developed targeted solutions to deal with them. 
First, deploy intelligent monitoring system. When the software automatically detects that the system is not operating properly, it automatically triggers the next step of fault analysis to locate the root cause and fix the fault rapidly. 
Intelligent monitoring can be performed in either direct or indirect mode. In a direct mode, the software monitors the indexes of key facilities such as the environment, hardware (compute, storage and network) and cloud OS. Once failure data occurs, it directly alarms and locates the fault. In an indirect mode, the software monitors the KPIs of 5G services and makes a multi-dimensional comparative analysis. Through the analysis, it finds abnormal KPIs and determines whether a fault has occurred, which triggers the further fault correlation analysis and location. The multi-dimensional comparative analysis can be made from several aspects. 
Second, introduce automatic operations to analyze the root cause of a fault through the vertical and horizontal correlations. That is, after the monitoring system detects a system failure, it triggers hierarchical alarms and log correlation analysis in the vertical direction, and also correlation analysis of intra-NF and inter-NF micro-services at the horizontal level. 
In the vertical architecture, network architecture includes three layers: physical layer, virtual layer, and service layer. When a fault occurs at the bottom layer, upper-layer services will be affected. The key to vertical correlation is to identify where the root cause of the fault is located. Is it at the PIM layer, VIM layer, or VNF service itself? 
At the horizontal level, there is service flow correlation and mutual influence among micro-services within a VNF. The horizontal correlation between micro-services can help to find the root cause of a problem. The key to horizontal correlation is to associate one faulty NE with another service-related faulty NE, associate one faulty micro-service with another micro-service that really goes wrong, and identify the really faulty micro-service or component on the same application layer. 
In addition to fault monitoring and hierarchical correlation analysis of fault root causes, other key technologies such as fault self-healing, global perspective, cross-domain data collection, network topology management, one-click automatic tests, and one-click automatic service deployment are all mature commercial capabilities for intelligent operations. 

Network Slicing to Bring Value-Added Services in Network Operations  
Network operations give rise to network slicing, which allows various industries to create more applications based on the 5G network. As a result, abundant vertical industry applications such as automatic driving, industrial control, smart grid, big video, and AR/VR will be widely used in our daily life. Ultra-large bandwidth, ultra-low latency, and massive connections will be ubiquitous in the near future. This is not only a major challenge to 5G core network operations, but also an important opportunity for the prosperity of 5G core networks. 
5G core network is essentially a virtual network that has all features of virtualization. Network slicing is an important feature. Through flexible allocation of network resources and flexible combination of capabilities, logical subnets with different features are virtualized based on a physical network to meet the customization needs in different scenarios.
Operations based on network slicing in essence are to provide the whole life-cycle management of slice instances, including their design, commissioning, SLA guarantee, and termination. While bringing great flexibility, network slices also increase the complexity of operation and management. It is an inevitable trend to enhance the automatic slice management capability based on artificial intelligence (AI). 
After the introduction of AI, the slice management system can automatically implement management policies according to the decisions output by the AI training platform, and give the network the capabilities of intelligent perception, modeling, commissioning, analysis, judgment and prediction, so as to achieve a perfect balance between slice flexibility and management complexity. Through intelligent slice commissioning, intelligent slice SLA guarantee and intelligent slice closed-loop operations, end-to-end slice delivery and management can be realized. It can be predicted that the combination of AI and 5G slicing networks will promote rapid network development and evolution.
ZTE will reduce the complexity of system maintenance brought about by the 5G cloud native and service-based software architecture through intelligent delivery and operations in the construction of 5G core networks, helping operators worldwide build high-quality 5G networks, focus on 5G services, and create greater value for their customers.