A analysis team at Xi’an Jiaotong-Liverpool University (XJTLU) has drawn a roadmap to realising clever 6G communications, publishing an in depth framework for its Channel Foundation Model (CFM).
Unlike 5G, which depends on predefined guidelines and engineering optimisation, 6G wants stronger environmental consciousness, job adaptation, and autonomous optimisation. However, typical supervised-learning AI fashions will wrestle right here due to their dependence on manually labelled information, main to poor generalisation throughout situations and duties.
Introduction to the GREAT-X simulation platform
To tackle this, Professor Shugong Xu and his team at XJTLU’s School of Advanced Technology first proposed the CFM idea in 2025. Now, Professor Xu and co-author Jun Jiang, a PhD pupil, have laid out a technical pathway in a paper printed within the ZTE Technology Journal.
CFM treats the wi-fi channel because the core object of research, utilizing a “pre-training and fine-tuning” paradigm: fashions be taught general-purpose channel representations from giant, numerous datasets, then adapt to downstream duties.
Inputs can embody conventional channel information, resembling channel state info (CSI) and channel impulse response, and multimodal info like pictures and location information, enabling richer understanding of bodily obstructions, multipath reflections, and different environmental circumstances.
Plaque presentation ceremony for the Suzhou Low-Altitude Economy Technology Innovation Centre
At the mannequin degree, utilizing pre-training to extract generalisable options from giant datasets reduces the necessity for manually labelled information. And in contrast to typical wi-fi AI fashions, sometimes designed for a single job, CFM goals to present a unified channel illustration framework that helps a number of duties and allows extra environment friendly switch, generalisation, and adaptation throughout situations.
“This shift from AI-assisted communications to ‘AI native to the channel’ will help communication systems move beyond passively adapting to their environment to actively understanding and learning from it and optimising themselves accordingly,” Professor Xu explains.
Unveiling of the Low-Altitude Technology Joint Transformation Centre
From simulation to actuality
The team can be transferring CFM in the direction of real-world use, drawing on its strengths in industrial and tutorial collaborations to advance technical validation and know-how switch.
Through the XJTLU Low-Altitude Innovation Research Institute, researchers on the University are working intently with firms together with Suzhou Aviation Industry Group and Suzhou Low-Altitude Tech, and main analysis establishments just like the Gusu Lab and Suzhou Institute of Nano-tech and Nano-bionics, Chinese Academy of Sciences.
Professor Shugong Xu delivers a chat on Native AI and CFM by way of the Teaspark platform
Together, they’ve established the Low-Altitude Technology Joint Transformation Centre, which focuses on core 6G areas, and are offering a powerful industrial basis for deploying CFM in low-altitude situations.
The team has additionally autonomously developed the GREAT-X platform to simulate and take a look at CFM throughout complicated environments together with dense city areas, open suburban landscapes, and low-altitude flight situations.
Meanwhile, two associated initiatives, CSI-CLIP and CSI-MAE, have been open-sourced to assist international analysis in clever 6G communications, because the team works in the direction of integrating CFM into worldwide requirements.
Text and pictures courtesy of XJTLU Low-Altitude Innovation Research Institute and Research Engagement and Innovation Office
Translated by Xiangyin Han
Edited by employees editor and Xinmin Han