Cranfield Team Attended and Presented at IEEE VTC 2025-Fall in Chengdu

Researchers from Cranfield University showcased pioneering 6G and AI research at the IEEE VTC 2025-Fall in Chengdu, China. Their presentations covered breakthroughs in edge AI orchestration, federated learning, and data-driven network stability, alongside a tutorial on trustworthy communications for connected intelligence systems. The event strengthened global collaboration and highlighted CHEDDAR’s leadership in shaping next-generation wireless technologies.

Researchers from Cranfield University (Dr. Yun Tang, Dr. Mengbang Zou, and Prof Weisi Guo) sponsored by the CHEDDAR project participated the IEEE Vehicular Technology Conference (VTC) 2025-Fall in Chengdu, China, presenting three cutting-edge papers and delivering a comprehensive tutorial on next-generation wireless communications.

End-to-End Edge AI Service Orchestration: Led by Dr. Yun Tang with co-authors Udhaya Chandhar Srinivasan, Benjamin James Scott, Dennis Kevogo, Obumneme Umealor, and Professor Weisi Guo, this work introduces an automation framework for designing and provisioning edge AI services, using Large Language Model (LLM)-powered agents. The framework translates user intents (one end) into orchestrated edge AI services (the other end)—a crucial step in commercializing the AI services in the 6G network.
Paper link: https://arxiv.org/pdf/2503.11933

Precision Planning in Federated Learning: Dr. Jinsheng Yuan, Dr. Yun Tang, and Professor Weisi Guo presented a novel approach to Mixed-Precision Over-the-Air Federated Learning (MP-OTA-FL). By incorporating RAG-based LLMs for dynamic client profiling, the research optimizes the balance between computational efficiency and model accuracy while improving user satisfaction and energy savings.
Paper link: https://arxiv.org/pdf/2503.15569

Ensuring Network Stability Through Data-Driven Methods: Dr. Mengbang Zou, Professor Weisi Guo, and Dr. Yun Tang tackled the challenge of cascade stability in O-RAN load balancing. Their data-driven approach identifies dynamic models from real-time traffic data, ensuring stable network operations while preventing endless handover scenarios.
Paper link: https://arxiv.org/pdf/2504.04154

Knowledge Transfer Through Tutorial Delivery

Professor Weisi Guo also co-delivered Tutorial T14 on “Task-Oriented and Trustworthy Communications for Connected Intelligence Systems” alongside colleagues from Beijing Jiaotong University. The tutorial covered adaptive semantic communications, timeliness-oriented protocols, and trustworthy AI implementation in 6G networks—essential topics for the next generation of connected intelligence systems.

Dr. Yun Tang at the IEEE VTC 2025 Conference

Looking Ahead

These contributions highlight CHEDDAR’s commitment to addressing both fundamental and applied challenges in next-generation wireless networks. From AI service orchestration to network stability and trustworthy communications, our research is shaping the future of 6G technology.

Our Appearances

Dr. Yun Tang and Dr. Mengbang Zou attended the conference in person, engaging with the global research community and showcasing CHEDDAR’s 6G and AI research.
The event provided valuable opportunities for collaboration and knowledge exchange with international peers.

For more updates sign up to our newsletter https://cheddarhub.org/dynamics-integration-cloned/newsletter/

Go to Top