Neural-Network-Enhanced Holographic Metasurface With Superior Wavefront Manipulation Capability for Multitarget Wireless Power Transfer Applications

Neural-Network-Enhanced Holographic Metasurface With Superior Wavefront Manipulation Capability for Multitarget Wireless Power Transfer Applications

clockVolume: 74, Issue: 1, January 2026

authorMeijun Qu

Neural-Network-Enhanced Holographic Metasurface With Superior Wavefront Manipulation Capability for Multitarget Wireless Power Transfer Applications

Abstract
DOI: 10.1109/TMTT.2025.3615474   IEEEXplore: https://ieeexplore.ieee.org/document/11215863
Device
Neural-network-enhanced holographic 10.6λ x 10.6λ metasurface with wavefront manipulation capability for multitarget wireless power transfer (WPT)
Spectrum
20 GHz
Novelty
A joint algorithm based on high-resolution diffractive neural network (HR-DNN) and multifocus (MF) method optimizes the object wave. A de-emphasis method mitigates the local field distortion impacts. A quantitative evaluation method for surface wave propagation loss is developed to provide a comprehensive power distribution analysis technique for holographic metasurfaces.
Application
Multitarget wireless power transfer
Performance
18.91 dB measured peak signal-to-noise ratio (PSNR) ; 26.53% total WPT efficiency for up to 13 points with controllable intensity and shape.