Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

AI-Driven Design of Stacked Intelligent Metasurfaces for Software-Defined Radio Applications

Ivan Iudice, Giacinto Gelli, Donatella Darsena

TL;DR

This work addresses AI-native design for stacked intelligent metasurfaces (SIM) in SDR/NTN contexts by integrating a differentiable SIM model into NVIDIA Sionna to enable end-to-end optimization of precoding and metasurface configurations. A mathematically grounded SIM model is embedded as differentiable TensorFlow layers, enabling GPU-accelerated simulation and gradient-based learning within an end-to-end pipeline. Monte Carlo results show that data-driven SIM designs outperform model-based baselines and baselines without SIM, demonstrating the DoF benefits of wave-domain processing. The framework provides a scalable path toward hardware-in-the-loop experimentation and practical deployment of programmable metasurfaces in 6G/NTN systems.

Abstract

The integration of reconfigurable intelligent surfaces (RIS) into future wireless communication systems offers promising capabilities in dynamic environment shaping and spectrum efficiency. In this work, we present a consistent implementation of a stacked intelligent metasurface (SIM) model within the NVIDIA's AI-native framework Sionna for 6G physical layer research. Our implementation allows simulation and learning-based optimization of SIM-assisted communication channels in fully differentiable and GPU-accelerated environments, enabling end-to-end training for cognitive and software-defined radio (SDR) applications. We describe the architecture of the SIM model, including its integration into the TensorFlow-based pipeline, and showcase its use in closed-loop learning scenarios involving adaptive beamforming and dynamic reconfiguration. Benchmarking results are provided for various deployment scenarios, highlighting the model's effectiveness in enabling intelligent control and signal enhancement in non-terrestrial-network (NTN) propagation environments. This work demonstrates a scalable, modular approach for incorporating intelligent metasurfaces into modern AI-accelerated SDR systems and paves the way for future hardware-in-the-loop experiments.

AI-Driven Design of Stacked Intelligent Metasurfaces for Software-Defined Radio Applications

TL;DR

This work addresses AI-native design for stacked intelligent metasurfaces (SIM) in SDR/NTN contexts by integrating a differentiable SIM model into NVIDIA Sionna to enable end-to-end optimization of precoding and metasurface configurations. A mathematically grounded SIM model is embedded as differentiable TensorFlow layers, enabling GPU-accelerated simulation and gradient-based learning within an end-to-end pipeline. Monte Carlo results show that data-driven SIM designs outperform model-based baselines and baselines without SIM, demonstrating the DoF benefits of wave-domain processing. The framework provides a scalable path toward hardware-in-the-loop experimentation and practical deployment of programmable metasurfaces in 6G/NTN systems.

Abstract

The integration of reconfigurable intelligent surfaces (RIS) into future wireless communication systems offers promising capabilities in dynamic environment shaping and spectrum efficiency. In this work, we present a consistent implementation of a stacked intelligent metasurface (SIM) model within the NVIDIA's AI-native framework Sionna for 6G physical layer research. Our implementation allows simulation and learning-based optimization of SIM-assisted communication channels in fully differentiable and GPU-accelerated environments, enabling end-to-end training for cognitive and software-defined radio (SDR) applications. We describe the architecture of the SIM model, including its integration into the TensorFlow-based pipeline, and showcase its use in closed-loop learning scenarios involving adaptive beamforming and dynamic reconfiguration. Benchmarking results are provided for various deployment scenarios, highlighting the model's effectiveness in enabling intelligent control and signal enhancement in non-terrestrial-network (NTN) propagation environments. This work demonstrates a scalable, modular approach for incorporating intelligent metasurfaces into modern AI-accelerated SDR systems and paves the way for future hardware-in-the-loop experiments.
Paper Structure (7 sections, 20 equations, 3 figures)

This paper contains 7 sections, 20 equations, 3 figures.

Table of Contents

  1. Introduction
  2. Background and Related Work
  3. System Model
  4. Model-Based Optimization
  5. Implementation in TensorFlow
  6. Simulation Results
  7. Conclusion and Future Work

Figures (3)

  • Figure 1: SIM-aided downlink architecture with amplitude-controlled (red) and phase-controlled (blue) layers.
  • Figure 2: Bit-error rate versus energy contrast for QPSK modulation scheme.
  • Figure 3: Bit-error rate versus energy contrast for 16-QAM modulation scheme.