Research

New Architecture for Future Wireless Networks

To fully exploit the potential of network resources and achieve high spectral and energy efficiency, it is essential to develop intelligent and flexible network architectures for future communication systems. This research focuses on feedback-free transmission mechanisms, AI-based channel state information (CSI) prediction, and resource management challenges in sixth-generation (6G) and beyond networks, including cellular radio access networks (RANs), space-air-ground integrated networks (SAGIN), Internet of Vehicles (IoV), etc.

AI-Driven Network Traffic Prediction and Control

With the rapid evolution of AI technologies, data-driven approaches have become a key enabler for efficient and reliable network operations. To address the challenges in various network environments and resource constraints, this research leverage machine learning for network traffic forecast to predict dynamic traffic demands, and for traffic engineering to optimize traffic routing in large scale networks. The objective is to develop practical methods for operating network systems in a more scalable, robust, and efficient way.

AI for Resource Allocation and Resource Allocation for AI

Artificial intelligence is fundamentally transforming the way of optimization and control in next-generation wireless networks. This research focuses on the development of advanced radio resource management and scheduling algorithms leveraging deep learning and reinforcement learning. In parallel, it addresses the challenges associated with resource allocation for heterogeneous and distributed learning tasks, and leveraging over-the-air federated learning (OTA-FL) for achieving integrated computation and communication.

Autonomous Control and Integrity Assurance for Distributed Systems

To meet the heterogeneous service demands of future applications, unprecedented management agility and real-time adaptability are required for control in large-scale distributed systems. To address the physical constraints of wide-area deployments and control complexity and signaling overhead, this research try to unlock edge devices’ potential and harness their distributed intelligence for autonomous control, and at the same time, preserve integrity and ensure fairness in distributed autonomous systems.

Foundation Models for Communication and Networked Systems

Foundation models are opening new opportunities for intelligent design and management in communication and networked systems. This research focuses on the development of efficient and reliable large-model-based methods including novel reasoning mechanisms, knowledge-enhanced modeling, multi-agent collaboration, and the integration of models with different capacities, to be applied for assisting network design and planning, understanding complex network behaviors, and supporting self-healing and recovery from network failures.