Professor Ming Zeng from Laval University gave an academic report in Room 402, Mingli Building, Xianlin Campus, on June 8, 2026, at 13:30. Associate Dean Wanqi Yang of the School of Computer and Electronic Information hosted the report.
In the report, Professor Ming Zeng presented his team's latest research achievements in cutting-edge wireless and optical communication technologies. First, addressing the performance limitations of Over-the-Air Computation (AirComp) under channel state information uncertainty, he proposed integrating Fluid Antenna Arrays (FAA) into AirComp networks and utilizing the Block Coordinate Descent (BCD) method to jointly optimize resource allocation and antenna positions; experimental results demonstrate that this robust optimization scheme significantly reduces the system mean square error, offering a novel approach for future intelligent wireless network design. Second, targeting the bandwidth limitations and device nonlinearities faced by high-speed optical communication systems, he elaborated on a neural network-based Digital Predistortion (DPD) framework, highlighting the application of direct and indirect learning strategies along with deep learning models such as Convolutional Neural Networks and Bidirectional Recurrent Neural Networks (Bi-RNN). Experiments confirm that this scheme delivers superior compensation capabilities in severe nonlinear transmission environments, providing an effective solution for intelligent signal processing in next-generation high-speed optical communications.