Unleashing the Power of 5G Advanced and Wireless AI: Revolutionizing Cellular Networks for the Future

How 5G Advanced and AI-driven wireless systems are set to unlock the true potential of cellular networks, propelling us towards the era of 6G innovation.

As 5G technology continues to transform the wireless landscape, the integration of AI in future wireless systems is becoming increasingly critical. The complexity of 5G networks demands a paradigm shift, prompting the wireless industry to explore the vast possibilities of AI-driven design methods. With the advent of 5G Advanced, the role of wireless AI is expanding, introducing revolutionary applications that will redefine the way networks and devices are designed and operated over the next few years. Moreover, as we look forward to 6G, wireless AI is poised to become native and pervasive, fostering seamless communication between devices, networks, and across protocols and layers.

The Power of On-Device AI:

The implementation of AI in smartphones and other devices has been a game-changer in recent years. However, most AI applications were developed independently either on the device or within the network, hindering true end-to-end (E2E) systems optimization. On-device AI is now emerging as a game-changing technology for improving E2E optimization in 5G networks. It enables distributed processing across millions of devices, harnessing their combined computational power. Additionally, on-device AI allows for personalized AI model learning based on a user's data, without sharing sensitive information with the cloud. This not only improves network reliability but also addresses data sovereignty concerns.

The Rise of Cross-Node AI:

While on-device AI enhances E2E performance, true systems optimization requires cross-node AI implementation, where AI training and inference take place collaboratively across both the network and devices. This demands deep domain knowledge of wireless systems and AI expertise. The upcoming 5G Advanced, as defined in 3GPP's Release 18, focuses on cross-node AI to address specific challenges, such as dynamically adapting the Channel State Information (CSI) feedback mechanism, intelligent beam management, and enhanced positioning accuracy.

Transforming Communication with AI-Based CSI Feedback:

Traditional model-based CSI feedback techniques can lead to significant signaling overhead in massive MIMO radios, affecting network capacity and device battery life. An AI-driven approach for channel estimation, on the other hand, dynamically learns from the environment, improving link performance and accuracy, especially at the cell's edges. Qualcomm's sequential training technique enables this process without sharing data across vendors, ensuring efficient implementation.

Intelligent Beam Management with AI Algorithms:

AI is also revolutionizing beam prediction on millimeter wave radios. Instead of continuously measuring all beams, AI-based algorithms intelligently select the most appropriate beams when needed, predicting future beams without continuous measurements. This approach optimizes network throughput and reduces power consumption, significantly enhancing the performance of AI-based systems.

Precision Positioning with AI:

Another application of AI in wireless systems is precise positioning. AI-based multi-cell roundtrip (RTT) and angle-of-arrival (AoA) positioning techniques have been demonstrated in both indoor and outdoor environments. Additionally, AI-assisted RF fingerprinting can overcome non-line-of-sight channel conditions in industrial private networks, offering accurate and reliable positioning.

AI-Native 6G Air Interface:

To accommodate the demands of 6G, AI will play an even more prominent role in wireless system design. The 6G air interface must be AI native from the outset, adopting a data-driven design approach across all protocols and layers. This AI-native air interface will offer continuous performance improvements, dynamically adapting both the network and device sides based on local conditions. It will enable wireless networks to be dynamically customized for various deployment scenarios, radio environments, and niche use cases, empowering operators to deliver seamless and tailored communication services.

Conclusion:

The convergence of 5G Advanced and wireless AI marks a significant milestone in the evolution of cellular networks. With AI's ever-expanding role in system design, 6G is set to revolutionize wireless communication by offering higher data rates, improved capacity, and a diverse range of network capabilities. As the wireless industry embraces the AI-native approach, we are on the cusp of an exciting era where cellular networks will truly unleash their full potential, empowering us to connect, communicate, and innovate like never before.