Hybrid RIS-Aided Digital Over-the-Air Computing for Edge AI Inference: Joint Feature Quantization and Active-Passive Beamforming Design

TL;DR

The paper tackles the bottleneck of edge AI inference by enabling rapid, distributed feature aggregation via a Digital AirComp framework enhanced with a Hybrid RIS. It introduces HRD-AirComp, which uses Grassmannian-vector quantization to map high-dimensional features into discrete codewords and digitally modulates them for over-the-air aggregation, while a task-oriented surrogate function G guides joint optimization of bit allocation, transmission coefficients, EN receive beamforming, and hybrid RIS reflection. A four-block alternating optimization algorithm (JQAPB) achieves efficient, near-optimal solutions, with closed-form updates for RIS parameters that significantly reduce computational complexity over SDR-based methods. Experimental results on linear classification and multi-view object recognition show substantial gains in inference accuracy (up to ~24.7%) over baselines and close proximity to ideal feature aggregation, demonstrating the practical impact of integrating digital AirComp with hybrid RIS for scalable, edge-focused AI tasks.

Abstract

The vision of 6G networks aims to enable edge inference by leveraging ubiquitously deployed artificial intelligence (AI) models, facilitating intelligent environmental perception for a wide range of applications. A critical operation in edge inference is for an edge node (EN) to aggregate multi-view sensory features extracted by distributed agents, thereby boosting perception accuracy. Over-the-air computing (AirComp) emerges as a promising technique for rapid feature aggregation by exploiting the waveform superposition property of analog-modulated signals, which is, however, incompatible with existing digital communication systems. Meanwhile, hybrid reconfigurable intelligent surface (RIS), a novel RIS architecture capable of simultaneous signal amplification and reflection, exhibits potential for enhancing AirComp. Therefore, this paper proposes a Hybrid RIS-aided Digital AirComp (HRD-AirComp) scheme, which employs vector quantization to map high-dimensional features into discrete codewords that are digitally modulated into symbols for wireless transmission. By judiciously adjusting the AirComp transceivers and hybrid RIS reflection to control signal superposition across agents, the EN can estimate the aggregated features from the received signals. To endow HRD-AirComp with a task-oriented design principle, we derive a surrogate function for inference accuracy that characterizes the impact of feature quantization and over-the-air aggregation. Based on this surrogate, we formulate an optimization problem targeting inference accuracy maximization, and develop an efficient algorithm to jointly optimize the quantization bit allocation, agent transmission coefficients, EN receiving beamforming, and hybrid RIS reflection beamforming. Experimental results demonstrate that the proposed HRD-AirComp outperforms baselines in terms of both inference accuracy and uncertainty.

Figures (9)

• Figure 1: Illustration of the edge inference system with a hybrid RIS.
• Figure 2: The diagram of HRD-AirComp scheme.
• Figure 3: Convergence of the proposed JQAPB algorithm.
• Figure 4:
• Figure 7: Different inference performance metrics versus number of agents $K$.