Integrating Stacked Intelligent Metasurfaces and Power Control for Dynamic Edge Inference via Over-The-Air Neural Networks

2025-09-23

AI Summary

Conventional edge inference frameworks model wireless channels simplistically as additive noise, leading to high communication and computation overhead. Method: This paper proposes an end-to-end over-the-air neural network framework leveraging stacked intelligent metasurfaces (SIMs) and dynamic power control. It explicitly models the wireless channel as a trainable neural network layer; SIMs enable channel-state-aware signal processing and spatial-domain computation, bypassing traditional symbol-level channel estimation. Joint optimization of SIM reflection coefficients and transmit power achieves tight co-design of propagation, computation, and communication. Contribution/Results: This work is the first to deeply integrate stacked SIMs with end-to-end over-the-air computation, enabling the physical channel to directly participate in model inference. Experiments across diverse scenarios demonstrate significant energy efficiency gains—average power reduction of 37.2%—while preserving classification accuracy, establishing a new paradigm for high-efficiency, low-overhead dynamic edge inference.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Search and Optimization: Sampling/Simulation-based Search
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

This paper introduces a novel framework for Edge Inference (EI) that bypasses the conventional practice of treating the wireless channel as noise. We utilize Stacked Intelligent Metasurfaces (SIMs) to control wireless propagation, enabling the channel itself to perform over-the-air computation. This eliminates the need for symbol estimation at the receiver, significantly reducing computational and communication overhead. Our approach models the transmitter-channel-receiver system as an end-to-end Deep Neural Network (DNN) where the response of the SIM elements are trainable parameters. To address channel variability, we incorporate a dedicated DNN module responsible for dynamically adjusting transmission power leveraging user location information. Our performance evaluations showcase that the proposed metasurfaces-integrated DNN framework with deep SIM architectures are capable of balancing classification accuracy and power consumption under diverse scenarios, offering significant energy efficiency improvements.

Problem

• Bypasses treating wireless channel as noise for edge inference
• Eliminates symbol estimation at receiver to reduce overhead
• Balances classification accuracy and power consumption dynamically

Innovation

• Stacked Intelligent Metasurfaces control wireless propagation
• Models transmitter-channel-receiver as end-to-end Deep Neural Network
• Dynamically adjusts transmission power using user location