As 5G gains traction around the world, the industry is seeking ways to support the growing complexity of networks and services. Wang Qiang, vice president of big data and service products at ZTE, says that autonomous AI-backed services like uSmartNet 2.0 are the way to go.
5G has already permeated into many societies and vertical industries around the world, including manufacturing, telemedicine, the automotive sector and smart homes. The almost endless possibilities create exciting prospects, but at the same time mean that networks must have considerable flexibility to meet both current and future needs.
The increased complexity of 5G is generated by factors including the billions of mobile and IoT devices going live on networks, the need to cater to many specialised industrial uses, and the multiple technologies and spectrum bands that operate via the latest mobile generation. There is also a call to provide systems and technologies that can address scenarios covering a service’s entire lifecycle.
“Compared with 4G networks, the complexity of 5G continues to increase,” says Wang Qiang, vice president of big data and service products at ZTE. “In addition, cloud-based software-defined networking and network functions virtualisation have changed traditional operations for telecoms players.”
Smartening up
To support future requirements, the creation of autonomous networks supported by AI and that can harness the benefits of big data is increasingly being seen by operators as the way to go. That approach, says Wang, can enable operators to provide so-called ‘Self X’ capabilities, comprising self-configuration, self-healing and self-optimisation of end-to-end networks – as well as reducing opex.
It can also deliver a ‘Zero X’ experience to customers – referring to a zero-wait, zero-touch and zero-trouble service.
But creating autonomous systems poses challenges. “5G, IoT and AI all rely on intelligent tools and systems to reduce manual intervention and provide the ultimate user experience,” says Wang. “However, building autonomous networks is a long-term process that requires multiple resources, as well as knowledge and experience.”
Nevertheless, ZTE believes it has some answers. Firstly, the vendor has applied state-of-the-art technologies such as intent-based network planning functions and digital-twin capabilities to areas including fault analysis, capacity prediction, network-performance optimisation and intelligent network slicing.
Secondly, ZTE is closely involved in research and standardisation for autonomous networks in conjunction with standards bodies like 3GPP, the TM Forum and the China Communications Standards Association.
Alongside this, ZTE has launched the second version of uSmartNet, a specific product to help operators move towards autonomous networks, after launching the first version in 2019.
‘Smart brain’
Wang describes uSmartNet 2.0 as a “smart brain” based on big data and AI technology, enabling agile, service-driven networks. “The service systematically focuses on ubiquitous connection, intelligent operations and maintenance, and agile operation in value scenarios,” he says.
The intent-driven offering enables 3D real-time twin mirroring that can dynamically predict trends in service development, while virtual-reality interaction supports simulation and optimisation, allowing operators to achieve zero-wait service provisioning. In trials, the accuracy of traffic prediction has reached 90% or more, says Wang.
On top of that, uSmartNet 2.0 enables automatic troubleshooting, boosting the experience for different types of users across the board, plus end-to-end evaluation and optimisation for voice-over-new-radio – or voice-over-5G – video calls. “Via end-to-end home broadband management technology, users with low levels of satisfaction can be easily identified, with an accuracy of over 90%,” adds Wang.
The service also enables low latency over 5G private networks, ramping up efficiency and quality in different verticals such as the steel and power industries.
Finally, uSmartNet 2.0 provides effective support for business development through integrated data analysis – with ZTE estimating that the service leads to a 15% conversion rate for telemarketing of mobile apps.
The power of AI
Wang sees huge promise in AI tools such as uSmartNet 2.0. “AI is a necessary technology for intelligent networks,” he says. “Seamless connection of such tools can lead to faster model development. It can also help to locate issues more quickly.”
And moving into the future, ZTE plans to continue exploring the opportunities for autonomous networks and intelligent services to boost automation, fixed-mobile convergence, AI-based innovation and cross-domain collaboration.
Wang says AI allows more efficient scheduling, algorithm development and energy-saving too – with more efficient power being a crucial consideration in today’s telecoms market.
By harnessing this type of automation, there is huge potential to proactively accelerate the digital transformation, says Wang. This is certainly an area in which ZTE is looking to lead the way, with the promise of more diversified, customised and personalised services for a wide variety of different uses.
“ZTE is promoting the deployment of highly autonomous networks for operators by improving existing networks, promoting industrial digitalisation and building smart systems,” he says. “I believe in this way, we can fully unleash the power of the digital economy.”
