The 5G era is coming to us in big strides. Rich and convenient network services have brought great changes to life, and different kinds of network services are constantly emerging. Such superior network conditions have greatly changed the personal user experience and brought new business development space to various industries. Disruptive innovations of traditional businesses have approached us.

In the 5G era, to achieve agile, automatic, and intelligent network operation and maintenance (O&M), this target has brought great challenges to network service providers. In recent years,  automation and intelligence has become a hot spot at MWC, global SDN/NFV conferences and other global top events in the industry. In 2018, the first technical forum with the theme of O&M automation was officially held in Madrid. Global communication technology experts gathered to discuss how to make O&M easy and comfortable under the current technology development trend. DevOps, Big Data, AI, Machine Learning and other technologies are also playing a role in this field, letting many dreams come true.

What can be done with respect to automation and intelligence?  
For service providers, the two mountains that they should face are nothing more than CAPEX (Capital Expenditure) and OPEX (Operating Expense). Only by continuously reducing these two indexes can the service providers be dominant and take the lead in the industry. This cliché is simple, but it faces many specific challenges. ZTE is committed to providing its users with the most convenient and practical O&M concepts and technologies to help users achieve significant reductions in CAPEX and OPEX. The following is an analysis of intelligent application cases in two scenarios.

Cross-cloud collaboration from telecom cloud to public cloud – reducing CAPEX
The diversity and complexity of the business inevitably leads to uncertainty in the business model. The daily, weekly and monthly business traffic and quality attributes are different. If you want to ensure the continuous and stable customer perception, in the initial stage of network construction,as a network service provider, you can only consider the construction according to the maximum capacity requirements. However, this solution inevitably leads to the idle state of resources for most of the time. It is unwise for the operator to build a network in advance and waste a lot of resources for a year without undue and irregular tidal traffic.

The end-to-end cross-cloud orchestration solution provided by ZTE has solved this problem well. In the initial stage of network construction, operators can build networks based on the basic average size, and flexibly and dynamically adjust resource scheduling in the data center to meet the business implementation in most scenarios. When the telecom cloud data center is unable to carry the corresponding service requests under certain specific requirements, ZTE's cross-cloud orchestration system CloudStudio can realize the coordinated scheduling from the telecom cloud to the public cloud, temporarily apply for public cloud resources to carry the corresponding businesses, according to the customer's business demands, and only need to pay short-term rental fees to the public cloud platform to complete the expansion of business scale, greatly reducing the operator's CAPEX.

End-to-end root cause analysis – reducing OPEX  
With the increasing complexity of the network structure, system nodes and interfaces are gradually increasing. The biggest challenge of O&M comes from quickly locating faults after problems occur in these nodes. From the perspective of network O&M, the stability of the network is the cornerstone. Whether the cornerstone is firmly reflected in the network fault can achieve rapid positioning, and the important manifestation of the network intelligent value lies in the ability to quickly locate the network fault. According to the statistics of the existing network of a commercial office of ZTE, the average daily processing volume of an ordinary maintenance person is about 300 pieces of messages. This does not include a large number of invalid assignments and incorrectly assigned alarms, which will cost a lot of manpower and material resources.

ZTE's RCA model is based on the open source learning algorithm Apriori (a frequent item set algorithm for mining association rules) and Prefix-Span algorithm, which forms more than 8,000 alarm-related rules, especially for the significant improvement of the recognition rate of accidental associated alarms. According to statistics, after machine learning is adopted, the overall alarm compression rate can reach more than 90%, the number of personnel handling alarms can be greatly reduced, and the efficiency of fault location can be significantly improved. At the same time, in addition to the root cause analysis of the alarm, ZTE's RCA algorithm can also monitor and analyze the network performance indicators and abnormal logs. Performance anomaly detection is applied to the cloud management network system, which can detect system anomalies in advance and destroy the faults in the rudimentary stage. Intelligent log analysis helps locate problems and is used to detect anomalies and warnings. With the application and feedback of the intelligent fault location function, the fault location method is also continuously expanded and optimized, and the fault location in the 5G network will be more intelligent.

Intelligence road in the future
The intelligent road of 5G network is a long evolution course from full manual, semi-automatic, fully automated, to intelligent mode. In this process, technologies such as virtualization, DevOps, big data, machine learning, and AI all support the evolution of operation and maintenance step by step. These advanced technologies must be applied in real-world scenarios so that the ideal can be realized. The needs of operators are eternal, for they need the continuous reduction of OPEX and CAPEX while fully guaranteeing the quality experience of end users. ZTE is committed to helping operators apply advanced intelligent technologies to their daily operation and maintenance. With the help of years of experience in the telecommunications industry, ZTE is providing intelligent and integrated operation and maintenance solutions and applications to more than 50 operators around the world.

Imagine that one day in the future, special needs such as “Building an enterprise dedicated line” and “holding a concert” can be instantly satisfied by a customized network that is automatically created. For intelligent O&M, we can do more. It is expected that the intimate embracement of operators and the majority of vertical industry users will provide a better business experience for end users.