Stacked Intelligent Metasurfaces-Based Joint Source-Channel Coding for Image Transmissions
IEEE Open Journal of the Communications Society · 2026 · 7 · 2825-2837
DOI: 10.1109/ojcoms.2026.3677119 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Open Journal of the Communications Society, 2026. Verfasst von Inhyeok Choi und 1 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Japan Science and Technology Agency (JST) ACT-X, Japan Society for the Promotion of Science (JSPS) KAKENHI.
Abstract (Originalsprache)
This article proposes a joint source–channel coding (JSCC) framework based on a stacked intelligent metasurface (SIM)-enabled electromagnetic neural network (EMNN) with residual connections for image reconstruction over wireless channels. A residual connection is embedded in the multi-layer SIM architecture to alleviate the vanishing gradient problem inherent in deep EMNNs. Building on this structure, we formulate a joint optimization problem that minimizes the reconstruction error and devise a training algorithm that jointly updates the amplitudes and phases of the elements in both the transmitter and receiver SIMs via error backpropagation grounded in a physical wave-propagation model. Simulation results on the MNIST dataset show that the proposed residual SIM architecture enables high-quality image reconstruction even under low signal-to-noise ratio (SNR) conditions and significantly outperforms a standard SIM design without residual connections, demonstrating enhanced robustness against channel noise.
Förderung
- Japan Science and Technology Agency (JST) ACT-X
- Japan Society for the Promotion of Science (JSPS) KAKENHI
Zitieren
APA
Choi, I., & Kojima, S. (2026). Stacked Intelligent Metasurfaces-Based Joint Source-Channel Coding for Image Transmissions. IEEE Open Journal of the Communications Society, 7, 2825-2837. https://doi.org/10.1109/ojcoms.2026.3677119
Vancouver
Choi I, Kojima S. Stacked Intelligent Metasurfaces-Based Joint Source-Channel Coding for Image Transmissions. IEEE Open Journal of the Communications Society. 2026;7:2825-2837. doi:10.1109/ojcoms.2026.3677119
BibTeX
@article{Choi2026,
author = {Choi, Inhyeok and Kojima, Shun},
title = {Stacked Intelligent Metasurfaces-Based Joint Source-Channel Coding for Image Transmissions},
journal = {IEEE Open Journal of the Communications Society},
year = {2026},
volume = {7},
pages = {2825--2837},
doi = {10.1109/ojcoms.2026.3677119}
}
EUVIMED