To overcome the complexities of NTNs, autonomous networks will be game-changing, facilitating greater control, cost-efficiency and performance across everything from minimising RF interference to optimising performance for fast moving user terminals. But reaching the level of autonomy needed to achieve this performance won’t come easy, as communications service providers (CSPs), vendors and technologists alike work towards much-desired Level 3+ autonomous networks. As illustrated in our recent whitepaper, 自律ネットワークに向けた競争において先行する, CC is drawing on the power of deep tech to develop the technology needed for new levels of network autonomy.

And what is deep tech? Simply put, it’s a bold mindset that harnesses radical science and engineering to achieve transformational innovation. It is deep tech that turns visions into realities. In this case, transforming the NTN market and unlocking a new era of connectivity.

But before we dive deeper into how we’re working towards Level 3+ network autonomy, let’s explore why it’s needed.

The benefits of autonomous NTNs

Firstly, what are we talking about when we say NTNs? In the context of this article, we’re talking about NTNs as an extension of terrestrial networks – wireless communication systems that extend beyond the Earth’s surface and into space. There are now thousands of satellites in low-earth orbit (LEO) with many more new networks planned. Some of these constellations use proprietary connectivity with dedicated User Terminals, but the shift to standards based connectivity using off-the-shelf user devices is accelerating fast. The 3D connectivity space will be further enhanced with High Altitude Platforms (HAPS) beaming 5G from the Stratosphere in the not-too-distant future.

Within the NTN space (no pun intended), operators are facing several challenges. One is the need to get more performance out of these networks. Traditional techniques are increasingly reaching their performance limits, prompting operators to find new ways to utilise the available spectrum to improve performance and reliability. Along with this technical challenge, there’s the commercial question: how can operators cut costs while still improving service?

When it comes to managing a network, whether they be ground-based (terrestrial) or airborne (non-terrestrial), integrating the network under a unified interface can streamline operations to enhance efficiency and ensure continuous, reliable service delivery.

The key to achieving this goal is to design a system with interoperability, scalability and adaptability in mind. This ensures the network can flexibly meet both current and future challenges – a particularly important advantage for NTNs given the cost of launch and how long assets can remain in service.

This is where autonomous networks utilising sophisticated AI become essential. Through AI, you can create high performance networks that are self-configuring, self-healing and self-optimising.

In short, AI allows for:

• Predictive analytics for network optimisation
• Dynamic routing and handover management
• Fault detection and self-healing
• Security enhancements
• Customisation and personalisation
• Harmonising standards and protocols

So, what does this come down to? Ultimately, unlike manually controlled systems which are more prone to error and are less responsive, an AI-driven autonomous network can dynamically ensure resources are being used optimally for greater cost-efficiency and performance while optimising user experience, no matter where the user is connecting from.

Now you know the benefits, let’s get into how CC are actively working towards creating this technology.

Use cases and active AI development at CC

We’re actively exploring AI-driven autonomous network development with a strategic, deep tech approach, integrating cutting-edge automation with advanced AI and ML (Machine Learning) models that enable networks to learn, adapt and optimise performance in real-time.

This technology will play a vital role in developing the practical solutions to realise your path to autonomy and the value autonomous networks can offer.

AI-based service quality prediction and estimation

Communication networks need to operate efficiently to allocate sufficient resources for each user and provide the best experience without allocating excess resources that would either increase costs or be detrimental to other users. That said, measuring the quality of user experience is not straightforward. Traditional metrics such as throughput, dropped calls or signal quality (“number of bars”) are crude metrics, and alternatives like interacting with applications the user is running or inspecting the user’s data are hard to implement and raise privacy concerns.

To address this need, we developed an AI agent that can accurately predict the user experience in real time, running within an O-RAN network implementation as an xApp. The agent operates on standard data within the RAN (Layers 1 to 3), requiring no modifications or extra interactions with the user’s device. When deployed as part of an autonomous network, this agent balances the resource usage against user experience in real-time, providing improved, personalised experiences direct to the customer.

This presents remarkable benefits across numerous scenarios, including for NTNs. For example, in a time-critical scenario such as during a disaster where terrestrial network comms may be disrupted, NTNs play a crucial role in emergency response coordination. Here the AI agent could adapt network resources to prioritise communications for emergency services, aiding in rescue and recovery operations that can save lives and minimise the impact of the disaster.

Sustainable network management

To reduce energy costs and meet sustainability goals, solutions are sorely needed to maximise energy efficiency. Mobile networks often provide service through multiple layers with a coverage layer remaining continually active to transmit always-on signalling information, although this can often be supplemented by a capacity layer to provide additional bandwidth.

CC has developed an AI agent to predict the loading of individual cells within a network. Using these predictions, a cell can automatically power down the capacity layer when the predicted throughput can be delivered through the coverage layer alone. This technology can significantly improve the energy-efficiency and sustainability of a network, predicting user density and typical usage patterns with no reduction in service for users. For example, it’s well known that satellite transmitters consume considerable energy. AI can predict when lower demand allows for certain transmitters to be powered down, extending the lifespan of expensive satellite equipment and reducing the energy needed to maintain service – a crucial need given the limited energy resources available in space-based operations.

Enhancing interference management

Autonomous networks can also drive better management of interference. Inter-cell interference (ICI) poses a significant threat to network performance, reducing effective capacity and compromising user experience.

Traditionally, techniques to mitigate ICI have centred on frequency and time domain solutions, ensuring signals for different users are transmitted at different times and/or frequencies. Yet the emergence of heterogenous networks (HetNets) demands a more holistic approach. In this type of network, where HAPs are used to fill gaps in coverage within a terrestrial network, the network organisation may be more dynamic than static and will need continuous reconfiguration.

We’re exploring how to minimise network self-interferences where AI can predict where people are or will be, then shifting power between cells as needed. Using simulation tools, we’ll combine our findings with existing reduction techniques such as intermodulation, frequency planning or neural networks to compare strengths and limitations of each available technique to integrate this technology into our existing HAPS and 5G developments.

This will be immensely useful given that AI can intelligently manage the interplay between HAPs and ground stations, adjusting frequencies, power levels and beam directions to minimise ICI. This ensures that both systems can coexist without degrading the overall network performance and providing reliable connectivity.

At CC, we help our customers deliver the true value and opportunity of autonomous NTNs through deep tech innovation. By utilising AI to intelligently manage their network’s resources, you can ensure strong and stable connections everywhere, to anyone. These are just a few examples of our work, and we’re working on many others to help our clients get ahead of the market. From our groundbreaking work with SmartSky to enable office-quality connectivity onboard business flights to our collaboration with Stratospheric Platforms Ltd (SPL) to beam 5G from the stratosphere, we have both the practical and strategic expertise to level up your network.

Reach out to discover more about how CC can work with you to bring high-quality coverage and services to even the most remote areas for your customers, breaking down barriers and connecting more people to the digital world.