Let's Not Limit Our Thinking When It Comes To AI I Telecoms

Artificial Intelligence is disrupting industries on a broad scale and telecoms is no exception. Tipped to be one of the hottest topics at this year's Mobile World Congress, we can expect AI to take centre stage in many telecoms conversations in 2024.

That being said, it's crucial for the sector to move beyond discussing how to build better, more intelligent customer service chatbots - and focus on alternative ways to leverage AI to deliver advanced connectivity solutions.

In fact, the UK ranks third in the world for research and innovation in AI, and our recent Market Research Insights Report points towards a growing momentum and significant investment and collaborative R&D&I activities in UK telecoms. As part of the £250 million Open Networks R&D Fund from the Department of Science, Innovation and Technology (DSIT), for example, an estimated £94.6 million is being spent on AI research projects.

To build on and utilise the country's advanced R&D capabilities, collaboration between academia and industry is needed to help identify opportunities in telecoms where we can build on our existing strengths in AI innovation and drive growth and success in the sector.

The current state of AI in telecoms

Telecoms networks are already complex enough to manage and will only become more so with the transition to 6G. Alongside this, the industry continues to face the headwinds of high operating costs, saturated markets and immense volumes of operational and customer data.

AI offers the ability to see patterns in data more quickly and more pervasively than human operators and make the right calls to increase performance, reconfigure in the face of outages, respond to traffic changes and generally improve the entire customer experience.

So far, operators are still trialling the technology in isolated parts of the network to gain the confidence they need before deploying network-wide. Progress is encouraging and no doubt more will be announced at MWC, but the caution is justified.

While the obvious applications of AI in telecoms are network-wide, at the other end of the length scale, AI is making an impact on one of the smallest but most important parts of the network – the base station transmitter power amplifier (PA). This is one of the most power-hungry parts of the network, and as there are so many base stations, it contributes materially to the operators' electricity bills.

The technical term for this is PA linearization. The more linear we can make our PAs the more efficient they are and the less electricity they waste as heat.

AI can be trained to recognise the characteristics of the amplifier, and the model can be used to infer the correct operating conditions for maximum efficiency. And we don't need to stop there, but we can also include the receiver and signal processing, completely revolutionising the digital radio techniques we've been using for decades.

It's heady stuff, maybe a bit further away, but is another example of the all-pervasiveness of AI and the importance of continuing to invest and innovate in telecoms.

Staying one step ahead

Despite the opportunities and benefits of AI, there are still several challenges for telecom companies to grapple with, from integration with legacy systems to a lack of in-house talent and the unstructured collection and categorisation of data. Convening the ecosystem to identify, explore and jointly address these key issues and challenges will be vital to realising the full potential of the technology in the future.

Recognising the need for closer collaboration between leading academic and industry experts, the UK Telecoms Innovation Network (UKTIN) has established its Artificial Intelligence Expert Working Group to address key challenges and help identify new opportunities where the UK can lead with AI in telecoms, creating an environment that encourages open communication and fosters innovation.

Ahead of MWC, the group will be publishing a Future Capability Paper that provides an in-depth exploration of both the challenges and opportunities around AI in telecoms.

The whole world is rightly excited about AI, but the reality is we've only made marginal progress when it comes to applications of the technology so far. Let's not limit ourselves when it comes to thinking about the full potential of AI in telecoms. At the same time, let's be mindful of the fact that we're still on the upswing of the hype cycle, or we risk setting ourselves up for disappointment.

While there's certainly a cultural conversation to be had about whether or not we'll lose our jobs to AI, the right response is to begin training and retraining now for a future where AI takes centre stage in telecoms. This is true not just for AI, but other emerging technologies such as Quantum. Businesses must start thinking about and preparing for their impact now to stay one step ahead of the competition.