yijun feng

Deep-learning-aided Metasurface Design and Manipulation: Meta-atom Optimization, On-demand Scattering Control and Self-adaptive Retroreflection

Recently, metasurfaces have attracted unprecedented research interests due to their powerful ability of manipulating electromagnetic (EM) wavefronts over a subwavelength thickness and demonstrating many exotic physics phenomena and novel devices. However, conventional metasurface designs usually take “bottom-up” procedures by carefully designing and optimizing the composing meta-atoms to achieve more and more sophisticated EM functionality demands through pre-knowledge and certain EM analysis, which sometimes lead to low efficiency, time-consuming full-wave EM simulations or even unsuccessful tries. Recently, machine learning (or deep-learning) methods such as artificial neural network (ANN), convolution neural network, and generative adversarial network have brought great opportunities in metasurface studies to obtain more flexible control of EM waves, such as inverse design of antenna, circular dichroism, absorber, and information metasurface. In this talk, I will report our recent studies on deep-learning-aided metasurface explorations that could provide “top-down” designing procedures, as well as inverse design and control methodology. Particularly, I will first showcase a deep-learning-aided method to achieve inverse design of dual-spin and dual-frequency metasurface capable of providing independent wavefront tailoring in multiple information channels. Secondly, I will exhibit that well-trained ANN can be utilized to inversely design a reconfigurable coding reflective metasurface to achieve on-demand EM wave scattering intensity efficiently and dynamically. Finally, I will introduce a pre-trained ANN for the direction of arrival (DOA) estimation, which achieves high precision by establishing the mapping between the directions of the incident signal and the amplitudes of multiple harmonics generated by a space-time-coding metasurface. Then the DOA information is used to dynamically modulate the phase gradient of the same coding metasurface accordingly to achieve self-adaptive retroreflection. All the examples have been validated by experimental tests on prototypes, demonstrating the powerful ability of the marriage of machine learning and metasurface in high efficient, multi-functional, and on-demand EM wave manipulations. This work was supported by National Natural Science Foundation of China (NSFC) under Grant NO. 62271243, 62071215, and partially supported by Jiangsu Provincial Key Research and Development Program (BE2023030789), and Jiangsu Provincial Key Laboratory of Advanced Manipulating Technique of Electromagnetic Wave.

Yijun FENG received the Ph.D. degrees from the Department of Electronic Science and Engineering, Nanjing University, Nanjing, China, in 1992. Since then, he has been a faculty member and is currently a Full Professor and the Deputy Dean of the School of Electronic Science and Engineering, Nanjing University. From September 1995 to July 1996, he was a Visiting Scientist with the Department of Physics, Technical University of Denmark. From August 2001 to August 2002, he was a Visiting Researcher with the University of California at Berkeley, USA. He has authored or coauthored over 200 peer-reviewed journal articles and over 180 referred international conference papers. His research interests include the electromagnetic metamaterials and their applications to microwave and photonic devices, electromagnetic wave theory, and novel microwave functional materials. He has conducted more than 20 scientific research projects including National 973, 863 Projects, the National Natural Science Foundation projects and the National Key Research and Development Program in China. Dr. Feng has received the 2010 Science and Technology Award (first grade) from Jiangsu Province, the 2021 Science and Technology Award (first grade) from Shanxi Province, China, and the 2022 Science and Technology Award (first grade) from China Institute of Communications. He has served as the General Co-Chair or Technical Program Co-Chair of several international conferences, including General Co-Chair for the 2018 IEEE International Workshop on Antenna Technology, 2024 IEEE Asia-Pacific Conference on Antennas and Propagation, and Technical Program Co-Chair for the 2013 International Symposium on Antennas and Propagation.