Metasurfaces-Enabled Wave Computing for Future Wireless Systems: Opportunities and Challenges

TL;DR

The paper tackles digital processor bottlenecks in future wireless networks by proposing wave computing with programmable metasurfaces as in-channel analog processors. It presents two paradigms—interactive and non-interactive—and surveys prototype demonstrations and wireless applications, including ISAC, wave-domain AI acceleration, and computational EM imaging, all realizable at the speed of light in the propagation medium. It identifies open challenges such as analog–digital integration, robustness, hardware design, and material technology, and outlines future directions like goal-oriented and semantic communications, photonics integration, non-local metasurfaces, and AI-driven programmability. The work highlights the potential for ultra-low-latency, energy-efficient processing that could redefine wireless system architectures by performing meaningful computations directly within the wireless channel.

Abstract

The next generations of wireless networks are envisioned to integrate communications, sensing, and computing into a unified platform, demanding ultra-high data rates, submillisecond latency, and unprecedented energy efficiency. However, conventional digital processors face limitations in scalability, cost, and power consumption that hinder this vision. Wave computing, enabled by programmable metasurfaces, offers an alternative paradigm according to which signal processing operations are implemented in the domain of the propagation of electromagnetic waves. This approach transforms metasurfaces from passive wavefront shapers into functional analog processors capable of executing tasks such as beamforming, sensing, imaging, and machine learning at the speed of light with minimal power consumption. This article provides an overview of metasurface-enabled wave computing, highlighting its fundamental principles and key application scenarios for future wireless systems, including integrated sensing and communications, artificial intelligence acceleration, over-the-air channel estimation, and computational electromagnetic imaging. Future research directions are outlined in response to the major open challenges of the technology, aiming to enable large-scale deployment of wave computing in practical wireless networks.

Figures (3)

• Figure 1: The main principle of wave computing with programmable metasurfaces: unlike digital signal processing (DSP), metasurfaces embed signal processing operations into their scattering profiles, transforming incident waves into processed outputs without need for digital conversion (top). Two paradigms of wave computing: interactive (bottom left), where sensing and feedback enable dynamic adaptation for tasks such as DOA estimation and beamforming; and non-interactive (bottom right), where predefined metasurface designs implement computations directly in the wave domain, such as edge detection or holography.
• Figure 2: Representative Prototypes of Metasurface-Enabled Wave Computing. Interactive paradigms are demonstrated using (Top left) a dual-functional STAR-RIS that enables simultaneous wireless communication and real-time DoA estimation within a single aperture. ( omam2025star_all), along with (Bottom left) a neuromorphic metasurface capable of task-driven analog inference ( mogh2024). (Top right) Non-interactive paradigms include a DMA-based platform for computational EM imaging ( 10564005_all) and (Bottom right) a D²NN performing fixed-function optical classification and analog computation ( lin2018all_all).
• Figure 3: Wave computing in next-generation wireless systems, showcasing its application across diverse domains. (Top left) Wave-Computing-Enabled ISAC uses STAR-RIS to jointly serve communication and sensing by embedding processing into wave propagation. (Top right) Wave-Domain-Based AI employs neuromorphic computing metasurfaces, where nanostructured layers physically implement neural network computations for real-time inference. (Bottom left) Over-the-Air Channel Estimation leverages coded metasurfaces for compressed sensing and DOA estimation. (Bottom right) Computational EM Imaging replaces conventional raster-scanning with wave-based coding to reconstruct images using random scattering, enabling faster and more efficient imaging.