By Wang Qiang, Vice President, ZTE
Autonomous networks (AN) will empower and accelerate the digital transformation of telecom operators and other industries. Under the trend of industrial digitalization, communication service providers (CSPs) are moving faster to build highly intelligent, automated cloud networks. Thus, ZTE has also been vigorously exploring for years in terms of three key aspects.
Three Aspects Leading to AN Concept
The first aspect is business-orientedness. To meet the service requirements, autonomous domains are coordinated to facilitate network transformation, thereby supporting business operations.
The second aspect is openness. Fulfillment of service requirements relies on end-to-end, cross-domain coordination, which is supported by AI-native capabilities within each domain. The ability to share capabilities and data and stay open for coordination enables closed-loop automation from process design to business operations. ZTE APIs applied to AN use cases comply and pave the way for a leading open network architecture in the industry.
Finally, the third aspect is network performance. All the efforts are made to fulfill the goal of "Zero-X services and Self-X capabilities". To this end, network performance plays a key role. In the scenarios of production and operation support, ZTE assists CSPs in evaluating autonomous capabilities, to see whether the network performance meets the requirements. Such efforts will help achieve an advanced level of autonomy, and in turn contribute to the criteria for level defining.
uSmartNet2.0: Addressing Pain Points for Telecom Operators
How can these three aspects be implemented? In this regard, ZTE has developed uSmartNet2.0 AN solution and putting it into extensive commercial use. Currently, ZTE has teamed up with more than 80 partners, and built over 100 sites globally.
As for industry standards, ZTE held the plenary meeting of CCSA TC7 last year. Holding 27 key positions in global standards organizations, ZTE plays an active role in AN standards formulation, such as requirements for capabilities at different levels, technology evolution, and evaluation system. Together with partners, ZTE has worked out 3 3GPP standards, 8 CCSA standards, and over 10 industry specifications and improvement solutions for different scenarios. In terms of awards granted by the TM Forum, ZTE has won "Outstanding Use of TM Forum Assets" and "People's Choice" in 2022.
Nowadays, certain matters are capturing the attention of the entire industry. This includes ways to make user experience and network performance predictable and formulating and executing optimization strategies securely and efficiently. ZTE builds digital twin networks by using a sensing model for spatial-temporal service distribution. In this way, changes can be predicted in real-time, and the optimal strategy for all-domain coordination will be auto-generated. Based on the auto-sensed differences with the physical networks, digital twins will then self-optimize. In the first trial runs, over 90% of physical network coverage can be accurately reflected by the digital twin networks in real-time, and the accuracy of traffic prediction reaching 90%. In addition, with automatic detection and AI-based UAV survey, faulty sites can be identified five times faster than before.
At the heart of the uSmartNet solution is the capability to predict user experience and network performance and automatically formulate and execute optimization strategies securely and efficiently. ZTE’s solution uses a network digital twin for sensing and predicting spacio-temporal service distribution. Deviation from intended experience and quality levels can be predicted in real time, prompting the iterative adjustments of network configurations across domains in real time. First pilots indicate that over 90% of physical network coverage can be accurately reflected with the digital twin network in real-time, and traffic prediction accuracy exceeds 90%.
Additionally, automatic detection, particularLY when enhanced with complementary techniques such as an AI-based UAV survey, allows CSPS to spot and restore faulty sites five times faster than before.
Enhancing the user experience is always a top priority for CSPs. With uSmartNet2.0, CSPs can address this issue effectively. This solution offers intelligent and real-time identification, prediction using the sensing model, and automatic troubleshooting. Therefore, uSmartNet2.0 can significantly increase satisfaction among individual, household, and industrial users. More specifically, the platform's end-to-end VoNR service system enhances the quality of experience (QoE) of video calls, as measured by vMOS, which is one of the leading indicators in the industry. Also, via end-to-end home broadband management technology, users with low satisfaction can be easily identified with an accuracy of over 90%. In the steel and power industries, uSmartNet2.0 guarantees low latency of 5G private networks, enabling CSPs to provide better services for verticals.
This year, ZTE provides open capabilities for NEs and OMCs in different scenarios, fully supporting the level-3 AN. These open capabilities include a large amount of industry-first applications. In addition, OMCs are fully interconnected in personal service, CN and TN, establishing cross-domain and intra-domain closed loops for voice and data services.
AN Development: Creating Value through Practice
The development of AN must be coupled with practical experience in order to continuously realize its value in real-world applications. A large amount of data is generated in the operation support systems. Through integrated data analysis, ZTE has built a strong capability in user profiling, with world-class performance in the timeliness and accuracy of data analysis. In fact, with uSmartNet2.0, the telemarketing conversion rate for mobile apps rises to 15%, and the cross-promotion of video services on mobile apps and smart TVs is realized for the first time. At present, ZTE's user profiling technology has been used by more than 500 enterprise customers.
ZTE has recently made significant strides in customer developments related to AN. One such example is the company's collaboration with China Mobile to construct quality centers, which involved implementing more than 30 metrics and processes to L4. ZTE also deployed its AN solution for Telkomsel in Indonesia, enabling the achievement of digital and intelligent network management. Furthermore, the company is continuously working to improve previous projects, such as those in Thailand and Uganda, with a focus on enhancing network quality and customer experience through AN.
In Thailand, AIS and ZTE have jointly launched a joint “A-to-Z” innovation center. This center serves as a hub for research and application, aiming to build highly autonomous networks for all scenarios. ZTE will continue to work with partners to deploy L3 AN and seek breakthroughs in L4 capabilities.