logo_header
  • Topics
  • Research & Analysis
  • Features & Opinion
  • Webinars & Podcasts
  • Videos
  • Dtw

China Mobile aims for widespread network autonomy by 2025

Dawn BushausDawn Bushaus
10 Aug 2023
China Mobile aims for widespread network autonomy by 2025

China Mobile aims for widespread network autonomy by 2025

China Mobile’s Dr. Lingli Deng believes her company is more ambitious than other service providers in its quest to achieve highly autonomous networks (ANs). “We are targeting Level 4 by 2025 for almost all the services that we are currently providing and that we are expecting to deliver in the coming years – it is the largest-scale production autonomous network practice ever,” she says.

Network automation levels 0-5

China Mobile is an active participant in TM Forum’s Autonomous Networks Project, which has developed a six-level AN maturity model (IG1252) that CSPs can use to measure their progress. Each level has a set of characteristics describing the evolutionary stage of a CSP’s journey from fully manual to fully autonomous operations. The model progresses from Level 0 to Level 5, with degrees of operations autonomy increasing at each step as shown in the graphic .

A survey conducted this spring and summer for TM Forum’s Digital Transformation Tracker (DTT 7) finds that about 20% of CSP respondents are using this model to track their progress. TM Forum is hoping to increase that percentage substantially over the next year.

Today, most telcos operate somewhere between Levels 1 and 3, but some such as China Mobile, China Telecom, China Unicom, MTN, Orange and Telecom Argentina have publicly announced their intentions to achieve Level 4 autonomy for at least some processes within the next two years.

Aiming for L4 across domains

Lingli Deng China Mobile

China Mobile is aiming for Level 4 autonomy for most processes across multiple domains including its mobile RAN and core, IP, optical and telco cloud networks. “The overall or average autonomous level for our production network today is 2.85,” says Deng, who works as a director researcher at the company and is active in standards development. “So, on average we have not achieved Level 3 yet, but this is what you would expect because we have 31 different branches, and each are operating in their own geographic area independently.”

Indeed, China Mobile’s networks and operations are vast, with 31 local operators serving 975 million mobile customers and 272 million cable broadband customers. China Mobile’s total investment in its 5G networks in 2022 alone was RMB 96 billion (about $13.5 billion). The company has deployed 1.28 million 5G base stations and serves 330 million 5G network customers.

China Mobile’s technology function is centralized, and it is using its end-to-end service operation layer as the catalyst to drive the overall development of its technical AN capabilities. “We have an end-to-end service operation management layer on the top of the five different technical domains, and we are specifying for each year the technical maturity level requirements,” says Deng.

At Levels 4 and 5, network systems leverage AI to acquire knowledge rather than needing to be programmed by humans through northbound interfaces. The systems learn to build knowledge from what they have observed and experienced in production.

Deng sees Level 5 as “very far reaching” and likely to incorporate technology such as generative AI. However, getting to that level may never happen in operators’ production networks because they will not be willing to exclude humans from the overall service delivery and assurance processes. “Perhaps Level 5 autonomy is only achievable from the technical perspective and not from the production network perspective – it will always be dependent on the service providers and their own policies,” she says.

Collaboration is needed

One of the biggest challenges China Mobile faces, according to Deng, is figuring out how to monetize its investments in technology through automation. “In addition to efficiencies from the internal operation perspective, we are also trying to set business value metrics, so that we can actually expose the automation and autonomy capabilities to our customers, or use those capabilities to find new customers,” she says.

China Mobile can’t solve this problem on its own, Deng adds. Industry collaboration is required. But collaboration isn’t easy when several layers of network management are evolving in parallel without clear consensus about whether to use a traditional network-centered approach to managing the transition to ANs or a “cross-layer approach”, as Deng describes it. In addition, there is a lack of consensus around the core reference architecture and APIs. For example, some industry efforts are focusing on introducing new APIs, while others promote enhancement of existing interfaces.

China Mobile, Orange, Vodafone and other large operators participating in TM Forum’s AN Project are embracing a cross-layer approach that relies on the Open Digital Architecture (ODA) and Open APIs. The ODA is a component-based approach that defines standardized, interoperable software components organized into loosely coupled domains. These components expose business services through Open APIs, which are built on a common data model.

“We have been quite dependent on organizations such as TM Forum to help us share our vision and requirements and get partners to work with us along the way,” says Deng. Those partners include not only providers of network elements, devices and network management systems, but also “fellow service providers who share the same vision, who share the same provider portfolio,” she adds.

“If we share the same vision, then we can actually combine our influence and push the vendors to move forward,” Deng explains. “There’s always controversy about how to achieve cross-layer autonomy. But it’s important for us to get a consensus as an industry so that different service providers won’t invest on their own in their specific ways.”

She adds: “It’s very much like the approach that we have been taking inside China Mobile. We have 31 different branches and they have been exploring on their own. But the resources they have to invest are quite limited. If they all go on their different ways, it will be a headache for us to push the target.

“It’s the same story for our industry. If we can get a consensus at the industry level among different service providers, it will be much easier, it will be much more economic, and it will be much more efficient to get to the goal.”