Implementing Neural Networks Over-the-Air via Reconfigurable Intelligent Surfaces

2025-08-03

AI Summary

To address the high communication and computational overhead in deploying neural networks for wireless edge intelligence, this paper proposes AirFC—a novel analog neural computing paradigm leveraging reconfigurable intelligent surface (RIS)-assisted MIMO over-the-air computation. AirFC uniquely exploits joint RIS–transmitter–receiver channel control to physically emulate fully connected layer weights via the wireless channel response, eliminating explicit channel estimation. By harnessing channel reciprocity, multi-RIS spatial diversity, and a semi-closed-form alternating optimization algorithm, it supports both centralized and distributed training. Simulation results on image classification demonstrate that AirFC achieves accuracy comparable to software-based implementations—without any digital computation at the edge devices—while significantly reducing terminal energy consumption and latency. These results validate AirFC’s feasibility and effectiveness for deploying deep neural networks in resource-constrained wireless edge environments.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Computer Vision: Learning & Optimization for CV
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

In this paper, we investigate reconfigurable intelligent surface (RIS)-aided multiple-input-multiple-output (MIMO) OAC systems designed to emulate the fully-connected (FC) layer of a neural network (NN) via analog OAC, where the RIS and the transceivers are jointly adjusted to engineer the ambient wireless propagation environment to emulate the weights of the target FC layer. We refer to this novel computational paradigm as AirFC. We first study the case in which the precoder, combiner, and RIS phase shift matrices are jointly optimized to minimize the mismatch between the OAC system and the target FC layer. To solve this non-convex optimization problem, we propose a low-complexity alternating optimization algorithm, where semi-closed-form/closed-form solutions for all optimization variables are derived. Next, we consider training of the system parameters using two distinct learning strategies, namely centralized training and distributed training. In the centralized training approach, training is performed at either the transmitter or the receiver, whichever possesses the channel state information (CSI), and the trained parameters are provided to the other terminal. In the distributed training approach, the transmitter and receiver iteratively update their parameters through back and forth transmissions by leveraging channel reciprocity, thereby avoiding CSI acquisition and significantly reducing computational complexity. Subsequently, we extend our analysis to a multi-RIS scenario by exploiting its spatial diversity gain to enhance the system performance. Simulation results show that the AirFC system realized by the RIS-aided MIMO configuration achieves satisfactory classification accuracy.

Problem

• Emulate neural network FC layer via RIS-aided MIMO OAC
• Optimize RIS phase shifts to minimize OAC-FC mismatch
• Train system parameters using centralized or distributed strategies

Innovation

• RIS-aided MIMO OAC emulates NN layers
• Low-complexity alternating optimization algorithm
• Centralized and distributed training strategies