Improve the quality of service on mobile networks

Mon 21 Jun 2021

The gains brought by AI functionalities will therefore have to be significant to justify the necessary investments.

Paul-Michel Bognier

What network engineer has never dreamt of having the right tools to predict and even solve a problem before it occurs? With AI, the dream might come true… It promises to provide operators that will know how to leverage it, the opportunity to significantly improve their quality of service and boost their competitive edge.

Quality of service: A fundamental challenge for telecom operators

To maintain its leadership and attractiveness, an operator must provide its customers with the best possible quality of service (QoS) on its mobile network, in particular: optimal coverage, accessibility and continuity throughout the operated territory, good voice quality, and convenient data rates.

Collection, measurement and analysis tools: Precious allies of network engineers

Since the emergence of mobile networks, operators have developed and evolved an ecosystem of tools that allows them to monitor the quality of service of their network on a daily basis.

Some commonly used methods include:

  • Collecting counters on mobile network equipment to continuously monitor the performance and health of the network;
  • Field measurements (Drive tests) and the collection of technical data from customers' smartphones (Crowdsourcing) provide an accurate view of the real customer experience;
  • Real-time collection and analysis of networks links traffic via probes installed at key points;
  • Self-organising networks (SON) technology that allows the auto-configuration, auto-exploitation and auto-optimisation of mobile networks equipment.

AI’s role in the quality of service

AI is still in the early stages in the field of quality of service, but it seems very promising. New generation SON (NG SON) tools based on 5G and machine learning are an illustrating example. Indeed, what network engineer wouldn't dream of having tools capable of accurately predicting the appearance of a quality of service problem in a zone, or even tools capable of automatically modifying network settings before the problem appears?

However, for these AI-enabled tools to be fully effective, the operator will need to collect, store and process even more technical data than they do now.

The gains brought by AI functionalities will therefore have to be significant to justify the necessary investments (infrastructure, equipment, software, increased skills of technical teams...).

AI or field of possibilities?

With the promise of AI, the ecosystem of tools for managing a mobile network and its quality of service will evolve significantly over the next five years. Tools that we don't even know exist yet will gradually be implemented. The job of network engineer will no longer be the same as it is today, and the "Network Data Scientist" component will become increasingly important.

The operator that can most effectively leverage AI to operate its mobile network will undoubtedly have an advantage over its competitors.

Extract from our white paper : Challenges and advancements in the era of data and artificial intelligence

Paul-Michel Bognier

Head of the End-to-End Network Performance Division