Level 4 Autonomous Networks Achieve Resilience, Optimization, and Agility in Live Operator Deployments

Autonomous networks have been discussed for more than a decade as telecoms operators strive to simplify, automate and optimize their networks. A step change in technology adoption is needed to create zero-touch, zero-trouble networks that can self-heal and ensure maximized uptime and availability to support the rapidly expanding portfolio of network-dependent services. Theoretical talk has helped frame autonomy but now we are seeing real action with operators adopting Level 4 autonomous networks. 

Level 4 attributes are set out in TM Forum’s Autonomous Network Level Model within which Level 5 describes fully autonomous operations. A Level 4 autonomous network can detect, predict and repair issues autonomously while focusing on self-healing and self-optimization in support of the customer experience. To achieve this the network needs to operate across multiple domains to resolve issues before they affect users. The aim is that the customer never becomes aware of network constraints and high-quality experiences are delivered seamlessly without interruption.

Among the operators already embracing the technology, South African operator MTN is utilising the Huawei Xinghe Intelligent Autonomous Driving Network to enable automation of network analysis and optimization tasks. These include automatic traffic suppression analysis and real-time visualisation, precise global path optimisation, predictive operations and the application of AI to simplify network operations. 

Driven to autonomy

MTN faces significant challenges from rotating power outages that can last for up to ten hours, causing routers and base stations to shut down and resulting in network congestion and packet loss. Manual analysis of power supply status requires substantial volumes of data to be combined, and traditional coding methods have proved inefficient and inaccurate. This leads to delayed response and a lack of flexibility to handle network issues.

By applying Huawei’s Network Digital Map and the built-in algorithm of the Autonomous Driving Network’s optimization capability, the MTN system can collect traffic information at the millisecond level. This is used to provide a heatmap to accurately restore the suppressed traffic distribution and visually identify network congestion in real-time. Visibility is also essential when it comes to precise global path optimization. If the network doesn’t have a real-time understanding of its topology, link status, bandwidth utilization, latency, and congestion points, it cannot be optimised. 

Therefore, continuous sensing and responding to dynamic network changes is essential to allow services to be steered along optimal paths that meet the SLA requirements of various service types. Significant benefits of Level 4 autonomous networks here include improved network resource utilization and enhanced operational agility.

The power to predict

Level 4 autonomous networks also support prediction of suppressed traffic through a characterized time-series model for base stations. This enables precise capacity extension suggestions for traffic growth, such as for large events. In addition, setting network SLAs in advance, including thresholds for packet loss or delays, powers the system to automatically design network paths to enhance service quality. MTN’s system offers more than 15 path calculation factors across bandwidth, latency and availability, among others.  

MTN has further simplified operations by working with Huawei to utilize AI to enhance quality, reliability and resilience. Next generation transport network technologies and agents boost network resource efficiency and cut down on device power consumption while AI agents support intelligent management and optimization of network devices, ensuring uninterrupted services and minimized environmental impacts from network operations.

A roadmap to success

Huawei launched its first digital twin for networks - Network Digital Map - three years ago and this enables operators to see not only the status of network devices but also the experience of network services that customers are receiving. The Network Digital Map has enabled a digital twin of the entire network to be built.

Orange is also utilizing autonomous network technology to transform its operation and maintenance activities. There is a need to simplify these and introduce a new model. The operator aims for its maintenance team to concentrate on more complex issues while AI copilots and agents support them by addressing more basic activities.

Another operator in the Orange group is MasOrange, which has the largest number of customers in the Spanish market. The operator deployed the first large-scale commercial network in Europe that operates autonomously with AI assistance in May 2025. With this deployment, MasOrange became the first European operator to deploy a Level 4 autonomous network. 

Automated autonomy

This network is already deployed in Alicante and Valencia in collaboration with Huawei and is supporting MasOrange’s aims of ensuring the network is secure, stable and flexible in complex environments while delivering an excellent experience to customers. Key features include the ability to detect and correct issues before the customer becomes aware of them.

The autonomous network can adjust latency parameters, speed and capacity in real-time to improve service quality even in areas where users are highly concentrated. The benefits include uninterrupted videos, more stable video calling, and lag-free gaming. The system enables each customer’s connectivity needs to be automatically adjusted based on their location and the content they want to access to ensure an optimised experience is delivered.

Disclaimer: Any views and/or opinions expressed in this post by individual authors or contributors are their personal views and/or opinions and do not necessarily reflect the views and/or opinions of Huawei Technologies.