Looking to the future, the convergence of 6G and artificial intelligence (AI) will revolutionize connectivity and technological advancements. This leap from outdated, hardware-centric, manually operated networks to intelligent, sensing, self-training, self-learning systems is made possible by the unparalleled capabilities of 6G and the steering power of AI. But can AI truly succeed without the foundation of 6G? Let’s explore the possibilities, hurdles, and possibilities that lie ahead at the intersection of 6G and AI.
6G Evolution Trends
Though 6G technology is still in its early stages, it is already taking shape with enhanced capacity, ultra-low latency, ultra-high reliability, and improved privacy. These improvements will facilitate faster data processing and real-time decision making, significantly improving the performance of AI applications. Edge computing is another trend to watch as it brings computational tasks closer to the data source, improving latency, privacy, and resource distribution across the network. The relevance of 6G to AI applications lies in its distributed computing network topology, cross-domain data fusion, and integrated sensing and communication. By integrating AI into mobile and edge networks, 6G will enable flexible AI model deployment and crowdsourcing of data from large-scale mobile and wireless environments.
6G enables new AI-based applications
Even before the 6G standard, we are already seeing glimpses of the potential of 6G networks. With 5G advancements, 6G will open up new frontiers for AI-based applications in these key areas:
- Data Efficiency: AI relies on efficient, low-latency data delivery to power its models. 6G will build a connectivity layer that efficiently feeds data to AI models, potentially creating an ecosystem for a variety of applications to thrive, including AR, VR, augmented reality, remote patient care, smart cities, and smart grids.
- Pervasive AI at the Edge: 6G is expected to extend AI beyond mega-scale data centers to the edge, leading to the emergence of AI-powered personal computers and devices. This expansion will pave the way for countless new applications, transform user experiences, and enable future technologies such as holographic communications and telepresence.
- Opportunities for Operators: 6G offers exciting opportunities for operators to integrate AI capabilities into their networks at the edge, enabling them to participate in a new set of AI applications as infrastructure providers. This convergence of AI and network infrastructure will not only revolutionize consumer applications but also enhance operational and management capabilities across industries.
- Data Privacy and Sovereignty: The rise of AI has brought data privacy and sovereignty to the forefront of concerns, creating opportunities for service providers to set up their own AI factories and leverage edge computing to address data privacy and security concerns, shaping the future landscape of AI applications.
The synergy between 6G and AI will enable a wave of transformative applications that will revolutionize user experiences, empower operators, and address critical data privacy and security challenges.
Key challenges for 6G to support AI
6G research is still in its early stages, and many challenges must be overcome to effectively support AI-based networks and applications. These challenges include:
- Mobile Computing Networks: The distributed topology of cloud and computing-based stations and mobiles poses great challenges for 6G networks.
- Data fusion and sensing: Integrating sensing and communication base stations to support sub-meter positioning of flying drones within a 1 km coverage area is a complex challenge that needs to be addressed.
- Validity and reliability: Ensuring efficient and reliable interactivity for a large number of AI robots is a significant challenge. Compared to 5G, improved user data rates, device density, reliability, and latency are essential to meet the demands.
- AI Native Network and Interface: Building a network that can support AI to be interconnected and work together across the network is crucial, including reducing network complexity and cost by allowing a single AI algorithm or model to control multiple functions and end-to-end operations of the network.
In addition to these technical challenges, data governance and security are key concerns for the efficient operation of AI within a network. Establishing a robust data governance framework is essential to sharing data across the network and enabling data capabilities while ensuring privacy, security, and regulatory compliance.
The sustainability imperative
The energy-intensive nature of AI, combined with the growing demand for AI-based applications, raises concerns about the potential for increased training costs during the transition to 6G. The urgency of addressing climate change is driving the need for increased energy efficiency in 6G. Fortunately, we don’t have to wait for 6G to start improving the energy efficiency of wireless applications. Open RAN architectures provide the opportunity to apply AI and ML in radio access networks, paving the way for sustainable practices and zero traffic.
Addressing the power consumption issues associated with training AI models requires advances in silicon and cooling technology, including the adoption of direct liquid cooling in large datacenters to significantly improve energy efficiency. Encouraging the use of open source models and implementing transfer learning can reduce the need to create and train many individual models and promote energy-efficient practices.
Streamlining model training and encouraging adherence to data schema and formatting standards is essential to reduce duplication and redundancy, and ultimately contribute to more efficient energy use. As 6G and 5G drive higher standards for integrated applications, there are opportunities to move towards a more unified approach, minimizing the need for multiple models and promoting sustainable energy practices.
Conclusion
AI is already revolutionizing industries and driving innovation across sectors, even without 6G. But to realize unprecedented advances in connectivity, we must be prepared to harness the full potential of AI and seamlessly integrate it with the upcoming 6G technology. While standards and specifications are still under development, working together to establish standards and define strategies is essential for successful integration of AI and 6G. Adopting AI now is not just an opportunity, it is a requirement and a gateway to unlocking the full potential of 6G and beyond.
