SYNCRO6G

Purpose and goal

SYNCRO6G aims to make over-the-air computation (OAC) reliable for real-world 6G edge networks with heterogeneous and imperfectly synchronized devices. The project will develop analytical models, adaptive AI-based synchronization control, and robust algorithms that mitigate timing and frequency misalignments, enabling low-latency, energy-efficient, and privacy-preserving federated learning over wireless networks at large scale.

Expected effects and result

The project will deliver synchronization-robust over-the-air computation methods for heterogeneous 6G edge devices, including analytical error bounds, AI-based adaptive synchronization control, and class-aware optimization algorithms. Expected effects include lower latency, reduced energy consumption, improved spectrum efficiency, and scalable privacy-preserving federated learning, strengthening Europe’s leadership in sustainable and AI-native wireless communication systems.

Planned approach and implementation

The project is organized into three technical work packages covering error modeling, AI-driven synchronization control, and robust optimization for heterogeneous devices, supported by training, management, and dissemination activities. Analytical modeling, deep reinforcement learning, and SDR-based experimental validation will be integrated to ensure both scientific rigor and practical feasibility. Open-source tools, datasets, and reproducible software will support broad impact and adoption.