Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Onboard Terrain Classification via Stacked Intelligent Metasurface-Diffractive Deep Neural Networks from SAR Level-0 Raw Data

Mengbing Liu, Xin Li, Jiancheng An, Chau Yuen

Abstract

This paper introduces a novel approach for real-time onboard terrain classification from Sentinel-1 (S1) level-0 raw In-phase/Quadrature (IQ) data, leveraging a Stacked Intelligent Metasurface (SIM) to perform inference directly in the analog wave domain. Unlike conventional digital deep neural networks, the proposed multi-layer Diffractive Deep Neural Network (D$^2$NN) setup implements automatic feature extraction as electromagnetic waves propagate through stacked metasurface layers. This design not only reduces reliance on expensive downlink bandwidth and high-power computing at terrestrial stations but also achieves performance levels around 90\% directly from the real raw IQ data, in terms of accuracy, precision, recall, and F1 Score. Our method therefore helps bridge the gap between next-generation remote sensing tasks and in-orbit processing needs, paving the way for computationally efficient remote sensing applications.

Onboard Terrain Classification via Stacked Intelligent Metasurface-Diffractive Deep Neural Networks from SAR Level-0 Raw Data

Abstract

This paper introduces a novel approach for real-time onboard terrain classification from Sentinel-1 (S1) level-0 raw In-phase/Quadrature (IQ) data, leveraging a Stacked Intelligent Metasurface (SIM) to perform inference directly in the analog wave domain. Unlike conventional digital deep neural networks, the proposed multi-layer Diffractive Deep Neural Network (DNN) setup implements automatic feature extraction as electromagnetic waves propagate through stacked metasurface layers. This design not only reduces reliance on expensive downlink bandwidth and high-power computing at terrestrial stations but also achieves performance levels around 90\% directly from the real raw IQ data, in terms of accuracy, precision, recall, and F1 Score. Our method therefore helps bridge the gap between next-generation remote sensing tasks and in-orbit processing needs, paving the way for computationally efficient remote sensing applications.

Paper Structure

This paper contains 20 sections, 5 equations, 3 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Methodology
  3. System Overview
  4. The Architecture of SIM-D$^2$NN
  5. Layer-by-Layer Diffractive Network.
  6. Wave Propagation Model.
  7. Inference from the Received Signal.
  8. Experiments and Discussion
  9. Data Generation and Preparation
  10. Quantitative Evaluation
  11. Main Comparisons
  12. Ablation Studies
  13. Discussion
  14. Conclusion
  15. Acknowledgment
  16. ...and 5 more sections

Figures (3)

  • Figure 1: High-level system overview. Comparison of processing progress: Traditional DNN-based learning from level-1 data vs. proposed D$^2$NN-based learning from level-0 data.
  • Figure 2: The illustration of the SIM structure and corresponding parameters.
  • Figure 3: Comparison of the visualization results under different methods.