In this year’s wireless automation roundtable during the 10th Mobile Broadband Forum in Zurich, operators and Huawei shared their experience and viewpoints on bringing automation capabilities to the network with AI technologies.

Stepping into the 5G era, mobile networks have a qualitative leap in key performance indicators (KPIs). These are the key capabilities allowing 5G networks support more diverse service scenarios and applications. However, 5G networks also bring the challenges of increasing CAPEX and OPEX for mobile operators. Moreover, service provisioning needs fewer manual configuration errors and quick service rollout. With 5G, it becomes essential to reduce the total cost of ownership of the networks, and at the same time to start to support more agile networks being able to enable many different types of services.

To deliver the 5G promise, we need more powerful control. The old OSS needs to transfer to a modern, autonomous network management environment using closed-loop control and AI techniques. Network automation is now becoming the indispensable 4th dimension of the 5G era together with eMBB, mMTC, and URLLC, and one of the most important driving forces for 5G service innovation and development.

AI technologies are playing a game changing role now. For network automation, AI is not only academic for mobile networks, it is also a necessary tool, like ABS in cars.

Changing the hype of AI and 5G to reality. Recently, in order to inspire more use cases of automation all over the world, GSMA released the whitepaper “AI in Network Use Case in China”. GSMA calls on the whole industry to work together to incubate more use cases and accelerate the digital transformation of the 5G era. In the white paper, GSMA released more than 30 cases in 7 categories, including network planning and construction, optimization and configuration, energy saving and efficiency improvement, operational services, maintenance and monitoring, service quality assurance and improvement, and network security protection. We can see that many operators, equipment vendors, and third-party vendors have already started to explore the development of automation use cases with AI technology applied, such as network traffic forecast, base station deployment automation, automatic fault location, and on-demand experience optimization. The paper also note that automation capability is vital for the first day 5G deployment, as it helps to meet the key challenges in the initial stage of cell planning and optimization.

Integrated automation and an AI framework is needed to help to tackle complexity. Scenario-based and personalized are the basic features of AI applications. In the future, the number of AI cases in telecom networks will be unimaginable. Integrated AI framework and standard workflow are necessary to effectively support massive AI cases and achieve sustainable development of network automation. What is more, as AI is enhancing RAN automation, we also need to apply AI to leverage reactivity, complexity and openness.

The industry need more cooperation and more open attitude to incubate various use cases, and so as to accelerate the implementation of autonomous driven networks (ADN). A network architecture featuring layered autonomy and vertical collaboration is required for the implementation of the intelligent autonomous networks. And realizing the closed-loop automation of the single domain is the foundation of the cross domain coordination. The whole industry need to work together on follow up work such as defining the scenario APIs to promote efficient collaboration between networks and processes. Many organizations in the industry like GSMA, ETIS, 3GPP, TMF are now working together on accelerating the network automation development and have made strong progress.

The integration of AI and communication networks will inject new vitality into mobile networks and open up unprecedented possibilities to bring the promise of AI and 5G to reality.