AI revolutionises error reduction for 5G & 6G networks

Researchers at Incheon National University in South Korea have developed an AI-powered method to improve the reliability and speed of 5G and 6G wireless networks, particularly for high-speed users.

This development simplifies the management of extensive signal data by using artificial intelligence to predict and mitigate errors, promising substantial improvements for applications such as high-speed travel, satellite communication, and disaster response connectivity. These advances come as mmWave technology, which employs high-frequency radio waves for data transmission, continues to be integral to next-generation wireless networks. This technology utilises massive MIMO systems, which involve large arrays of antennas working cooperatively.

Efficiently managing these complex antenna systems requires detailed knowledge of wireless environments, known as "channel state information" (CSI). This information can rapidly change due to user movement, resulting in the "channel aging effect," which can disrupt connections and cause errors. Addressing these challenges, a team led by Associate Professor Byungju Lee has introduced a new AI-centric approach known as "transformer-assisted parametric CSI feedback."

The technique zeroes in on critical signal parameters such as angles, delays, and signal strength, reducing the volume of information necessary for transmission back to the base station. Prof. Byungju Lee, explains, "To address the rapidly growing data demand in next-generation wireless networks, it is essential to leverage the abundant frequency resource in the mmWave bands. In mmWave systems, fast user movement makes this channel ageing a real problem."

The researchers employed a transformer model to understand and predict signal patterns, moving beyond previous methods like Convolutional Neural Networks (CNNs). Transformers can recognise short- and long-term changes in signal patterns, enabling real-time adjustments during user movement. The team reduced error rates by emphasising transmitting only the most crucial information—such as angles and delays.

Testing showed over a 3.5-decibel improvement in error reduction compared to conventional methods, with enhanced data reliability demonstrated using the bit error rate as a measure. The system was tested across a range of scenarios, from pedestrian speeds of 3 km/h to vehicle speeds of up to 60 km/h and high-speed environments like highways. Across all these scenarios, the new method consistently outperformed existing technologies.

Such advancements could enable uninterrupted internet for passengers on high-speed transportation, facilitate satellite communication in remote areas, and ensure connectivity during emergency situations when conventional networks might fail. Additionally, the method benefits emerging technologies, like vehicle-to-everything (V2X) communications and maritime networks. "Our method ensures precise beamforming, which allows signals to connect seamlessly with devices, even when users are in motion," states Prof. Lee.

This method introduces a new standard for wireless communication, ensuring the speed and reliability imperative for the next-generation networks.