Seamless LEO Constellation Collaboration - Tighter, Faster, Stronger

---

🔥 News 🔥

• [Dec 2025] ✨ SATITCH has been accepted to INFOCOM 2026!
• [Jul 2025] 🎉 SATITCH prototype officially released!

🌟 Overview

SATITCH is a cutting-edge framework that dramatically reduces the negative impacts of inter-SNO handovers and enables seamless collaboration between LEO constellations.

🎯 Key Achievements

• 75.8% reduction in flow-level disruption time
• Smooth real-time video streaming during constellation collaboration
• 94.57% improvement over state-of-the-art frameworks
• 30-second computation time vs. days for traditional methods

✨ Core Features

• 🌍 Nearly 100% Global Coverage

  • Multiple SNOs collaboratively provide network services for continuous global coverage

• 🤝 Seamless Constellation Collaboration

  • Flexible satellite resource allocation through RAN pooling layer

• 🚀 Near-Imperceptible Handover Latency

  • Proactive Authentication and Tunneling mechanism

• ⚡ Lightning-Fast Computation

  • Dynamic and Parallel SNO Association algorithm

🏗️ System Architecture

Component	Technology	Description
🛰️ LEO Simulator	StarPerf v2.0	High-fidelity satellite constellation simulation
📡 Mobile Network	Open5GS	5G core network emulation
📶 RAN Emulator	UERANSIM	User equipment and RAN emulation
🎥 WebRTC Server	Chrome WebRTC	Real-time video streaming

🚀 Quick Start

📋 Prerequisites

Before you begin, ensure your system meets these requirements:

Requirement	Version/Details
Operating System	Linux (Ubuntu 20.04+ recommended)
Python	3.10 or higher
Git	Latest version
Disk Space	~10 GB for datasets and outputs
RAM	8 GB minimum, 16 GB recommended

And ensure you have Conda installed.

conda create -n satitch python=3.10
conda activate satitch
pip install -r requirements.txt
# for further use, install gurobi from https://www.gurobi.com/academia/academic-program-and-licenses/

Step 1: Clone and Setup

# Clone with submodules for complete setup
git clone --recursive https://github.com/SpaceNetLab/Satitch.git
cd Satitch
# Conda environment setup
conda activate satitch

Step 2: Prepare Datasets

📊 Data Sources

Before generating datasets, familiarize yourself with our data sources:

• Satellite TLE Data: Celestrak
• Ground Station Data: Research Paper (arXiv)
• Crowdsourced GS Locations: Open Starlink GS

🛰️ Satellite Data Generation

cd SatelliteDataGen-Module
# Create directory for StarPerf .h5 output files
mkdir -p input_h5
# Place your StarPerf .h5 files in input_h5/ directory# Then run the generation scripts
python auto_generating.py
python select_timeslot_files.py
# Output will be in ./output_txt_integrated/

Copy generated satellite position data to the handover simulation directory:

# Copy these directories to inter-handoff-Starlink+Kuiper/
cp -r output_txt_integrated/STARLINK_KUIPER_POSITIONS_* ../inter-handoff-Starlink+Kuiper/

📁 Expected satellite data directories (click to expand)

STARLINK_KUIPER_POSITIONS_010/
STARLINK_KUIPER_POSITIONS_020/
STARLINK_KUIPER_POSITIONS_030/
STARLINK_KUIPER_POSITIONS_040/
STARLINK_KUIPER_POSITIONS_050/
STARLINK_KUIPER_POSITIONS_060/
STARLINK_KUIPER_POSITIONS_070/
STARLINK_KUIPER_POSITIONS_080/
STARLINK_KUIPER_POSITIONS_090/
STARLINK_KUIPER_POSITIONS_100/

📍 Ground Station (GS) Data Generation

cd ../GSDataGen-Module
# Choose your desired constellation alliance (e.g., "Starlink+Kuiper")cd Starlink+Kuiper
# Run batch generation script (one-click solution!)
python3 batch_generating_input_files.py

Copy generated GS files to the simulation input directory:

# Copy all generated test files
cp GS_TEST_*.txt ../../inter-handoff-Starlink+Kuiper/GS_INPUT/

📄 Generated GS data files (click to expand)

GS_TEST_CSGI.txt # For CSGI baseline
GS_TEST_SAFH.txt # For SAFH baseline
GS_TEST_SKYCASTLE.txt # For SkyCastle baseline
GS_TEST_SKYOCTOPUS.txt # For SkyOctopus baseline
GS_TEST_STARALLIANCE.txt # For StarAlliance baseline

Step 3: Run Handover Simulation

🔄 Handover Analysis

# Navigate to constellation-specific directorycd ../../inter-handoff-Starlink+Kuiper
# Run handover simulation (⚠️ This may take several hours)
python3 BatchProcessing.py
# Calculate disruption time metrics
python calculate_disrtuption_time.py

Note: Results will be saved in ./XXX_Handover/ directories

📊 Generate Visualizations

# Generate handover count visualization (Figure 7a in paper)
python draw_handoff_num.py
# Generate disruption time comparison (Figure 9a in paper)
python draw_disruption_time.py

Output plots will be available in draw_output/:

• [Evaluation]-Inter-SNO-Optimization-Starlink+Kuiper.pdf
• [Evaluation]-Disruption-Time-Reduction-Starlink+Kuiper.pdf

📚 Additional Resources

Constellation-Specific Guides

For detailed instructions on different constellation combinations:

Alliance Configuration	Documentation
Starlink + Kuiper	README
Telesat + OneWeb	README
Telesat + OneWeb + Qianfan	README

📊 Benchmarks & Performance

Performance Comparison

Metric	Baseline	SATITCH	Improvement
Flow Disruption Time	100%	24.2%	✅ 75.8% reduction
Handover Latency	100%	5.43%	✅ 94.57% improvement
Computation Time	48+ hours	~30 seconds	✅ 5760x faster

📈 Detailed benchmark results and experimental data coming soon!

🤝 Contributing

We welcome contributions from the community! SATITCH is actively developed at SpaceNet Lab - Tsinghua.

1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Open a Pull Request

For detailed guidelines, see our CONTRIBUTING.md

📖 Citation

If SATITCH helps your research or project, please cite our paper:

@INPROCEEDINGS{Lai2605:Seamless,
AUTHOR="Zeqi Lai and Yunan Hou and Boxuan Hu and Qian Wu and Jun Liu and Rui Chen",
TITLE="Seamless {Inter-Constellation} Sharing via {Handover-Aware} {Space-Ground} Association",
BOOKTITLE="IEEE INFOCOM 2026 - IEEE Conference on Computer Communications (INFOCOM 2026)"
}

📬 Contact & Support

• For technical questions and feature requests, please use GitHub Issues or Github PRs.
• For collaborations and partnerships, please contact us at:
  • zeqilai AT tsinghua.edu.cn
  • houyn24 AT mails.tsinghua.edu.cn
  • bxhu2004 AT stu.xjtu.edu.cn