Skip to main content

Harnessing the power of AI in autonomous RAN operations

Harnessing the power of AI in autonomous RAN operations

The business environment in the telecommunications industry is transforming and complexity is increasing fast, like in other industry sectors. Autonomous operations have become essential for managing that complexity cost-efficiently, and ensuring that subscribers have a smooth experience throughout the network.

At Nokia, we see artificial intelligence as a game changer for autonomous radio access network (RAN) operations. We are harnessing the power of AI in our Cognitive SON, a software feature of our market-leading self-organizing networks solution, Nokia MantaRay SON.

Before I go into more detail about Cognitive SON, let’s examine the concept of autonomous networks.

Artificial intelligence is essential for autonomous networks

Nokia is an active contributor to TM Forum, an industry alliance of over 800 companies across the world, including 10 of the top mobile network operators. TM Forum defines an autonomous network as “a system of networks and software platforms that are capable of sensing its environment and adapting its behavior accordingly with little or no human input”.

We have aligned the AI capabilities of Cognitive SON with the TM Forum’s autonomous network framework and together with other industry players, we are exploring use cases for AI-driven SON.

The framework defines a technical architecture that supports developers in designing and implementing autonomous solutions for network operations and optimization. It defines six levels of automation, with level 0 being fully manual and level 5 fully autonomous. The extent of human intervention and autonomous AI-driven operations are important in evaluating the network autonomy level.

I’m very proud that our Cognitive SON can reach TM Forum’s level 4 already today. This is thanks to the advanced artificial intelligence capabilities we have built into the solution. Level 4 represents highly autonomous operations – and you cannot reach levels 4 and 5 without AI.

Cognitive SON brings the next level of automation to network optimization

For our customers, typical optimization targets include performance, quality of service (QoS), planning, energy efficiency and operational efficiency. Over 120 customers worldwide have already chosen our MantaRay SON as the automation solution for the self-healing, self-configuration and self-optimization of their radio networks.

With Cognitive SON, we are helping our customers take the next, giant step in automated optimization. It uses established and enhanced AI algorithms to contextualize radio cells and orchestrate the SON modules autonomously without human intervention. This means that workflow execution becomes much faster and much more accurate.

Contextualization is essential for network optimization because each cell has multiple static and dynamic characteristics. The AI capabilities of Cognitive SON identify and set the dynamic context based on network KPIs and keep it always up-to-date.

Let me give you an example: it took less than 10 minutes for Cognitive SON to contextualize 70,000 radio cells in a large-scale customer network. Just imagine how long it would’ve taken to do the same manually! We are talking about several days of manual effort. In addition, it’s not even possible for a human engineer to identify all the dynamic characteristics of a radio cell, whereas it’s a simple task for AI-driven Cognitive SON.

Cognitive SON brings the benefits of AI-driven optimization also to multi-supplier environments and across technology domains, thanks to the open architecture and standardized interfaces of our MantaRay solution.

Nokia helps you unlock the power of autonomous RAN operations today

As the telecommunications industry starts to embrace digitalization in network operations, we will see the positive impacts on total cost of ownership (TCO). Today, operators spend a lot on manual operations and services, which could be almost completely automated.

Let me conclude by highlighting some impressive results from a Tier 1 operator’s network. This customer was able to enhance network performance KPIs by 30 percent, detect and repair network problems almost 90 percent faster and reduce operational effort by 99 percent.

Read our Cognitive SON Solution brief and visit the Nokia MantaRay SON webpage to learn how Nokia can help you today to start your journey towards autonomous RAN operations powered by artificial intelligence.

Sami Pohjonen

About Sami Pohjonen

Sami Pohjonen is the Head of MantaRay Product Management at Nokia, which encompasses radio and core Network Management, SON, and RIC. With over twenty-five years of international experience in the IT, software, and telecom industries, Sami has a proven track record in leadership and change management across start-ups, global consultation services, and product business environments. 

Connect with Sami on Linkedin

Article tags