日本av视频

The Wireless Networking and Mobile Computing Lab at 日本av视频 is dedicated to advancing research in wireless communication, mobile edge computing and AI-driven networking. Our lab focuses on next-generation networking technologies, including wireless federated learning, vehicular networks and intelligent control for IoT and mobile computing applications. Equipped with state-of-the-art software-defined radios, unmanned aerial vehicles and edge computing platforms, scalable, energy-efficient and AI-enhanced networking solutions can be developed. Through interdisciplinary research and industry collaborations, the Wireless Networking and Mobile Computing Lab aims to push the boundaries of wireless networking and mobile computing, solving real-world challenges in autonomous systems, smart cities and precision agriculture.

Research Area

• ML in wireless mobile networks
• Mobile terminal collaborations
• Cyber-physical systems

Research Project

Autonomous Vehicular Platooning in Smart Agriculture

The goal of this project is to realize timely and accurate control to enable coordinated and efficient operations for vehicular platoons with in the wireless federated learning framework.

To achieve this goal, we have designed:

1. An age-of-information-aware optimization framework
2. A fine-grained temporal learning architecture
3. A distributed ML algorithm for vehicle selection

Our preliminary results on two benchmark datasets show the effectiveness of our design.

Learning Accelerations for Unmanned Aerial Vehicles Trajectory Planning

The goal of this project is to characterize reconfigurable intelligent surface (RIS)-assisted unmanned aerial vehicle (UAV) communications and design responsive and accurate trajectory planning algorithms for the UAV based on computing acceleration techniques.

To achieve this goal, we have designed:

1. A new UAV-RIS-GT channel model under imperfect Channel State Information
2. A quadrotor UAV energy model for realistic trajectory planning
3. An adaptive UAV scheduling for joint optimization of trajectory and communication
4. An accelerated RL training via multithreading and federated learning

Our results demonstrate that our design achieves desirable speedup without compromising accuracy.

Learning-Over-the-Air: When Federated Learning meets Mobile Edge Computing

Issue. Ubiquitous Intelligence is the striking feature of NextG networks.

So What? It significantly impacts across multiple domains, including communications, industry, innovation, and even national security.

Problem. There is a significant gap between current AI and wireless technologies due to the widespread distribution of mobile devices and data.

Solution? FL in MEC (How FL is deployed/applied in MEC) and MEC for FL (How MEC platform supports FL) to achieve UI.

Benefits. Shaping and enhancing our day-to-day experience and the country's leading technological innovation worldwide.

Past Research Project

Mobile Edge Computing-assisted Task Offloading in the Internet of Vehicles