🚀 RFInject: Synthetic RF Interference Injection for Sentinel-1 SAR L0 Data

🔗 Dataset Links

📌 Motivation

• Radio Frequency Interference (\gls{RFI}) is a major source of performance degradation in modern Synthetic Aperture Radar (\gls{SAR}) missions.
• The Copernicus Sentinel-1 constellation is significantly affected, with numerous studies reporting its detrimental impact.
• However, the lack of standardized and reproducible datasets has so far limited systematic benchmarking of RFI detection and mitigation strategies.

---

🛠️ What RFInject Brings

RFInject introduces a methodology for controlled synthetic RFI injection into clean Sentinel-1 L0 raw bursts, enabling:

• ✅ Reproducible benchmarking of mitigation algorithms
• ✅ Realistic simulation while retaining authentic system properties
• ✅ Full parameter control over RFI characteristics

---

📐 Methodology Highlights

The framework is based on a parametric signal model:

• 🎯 Synthetic RFI generation by superimposing modulated chirp trains onto authentic Sentinel-1 radar echoes.
• 🧠 Spectral and statistical fidelity ensured to reflect real operational systems.
• 📊 Metadata-rich parameter sets controlling:
  • 📡 Waveform diversity
  • 🌍 Spatial extent
  • ⚡ Power scaling

---

📂 Dataset Features

• Clean Sentinel-1 L0 bursts → contaminated with controlled synthetic RFI
• Fully reproducible contamination scenarios
• Rich metadata for systematic testing across different algorithms and experimental setups
• Mirrored data access through the ESA Φ-lab bucket storage infrastructure, with public entry point through the ESA Φ-lab Hugging Face collection:
  🔗 https://huggingface.co/collections/ESA-philab/rfinject

---

🎯 Impact and Applications

The dataset empowers researchers to:

• 🕵️‍♂️ Detect RFI more reliably
• 🛡️ Mitigate its impact effectively
• 🤖 Develop learning-based solutions for robust RFI-resilient SAR processing pipelines