PaperlayerView exact revision on arXiv

Source-linked AI summary

Characterizing the Landscape of Open-Source Satellite Software

Jinfeng Wen, Qi Liang, Yuehan Sun, Federica Sarro, Ao Zhou, Xuanzhe Liu, Shangguang Wang

arXiv:2608.26211v1cs.SE

TL;DR

Existing research lacks a systematic empirical understanding of open-source satellite software. This paper analyzes GitHub projects to characterize ecosystem growth, software goals, and development practices, finding a growing but heterogeneous, task-specialized ecosystem.

  • Problem

    Existing studies lack a systematic understanding of the open-source satellite software landscape despite its growing role in satellite engineering.

  • Method

    The study mines 22,286 GitHub projects and manually inspects a representative sample of 646 projects to characterize popularity, software goals, and development practices.

  • Results

    The ecosystem shows sustained growth, 43 software-goal categories, and technological heterogeneity across 66 programming languages used in task-aligned ways.

  • Takeaways & Limitations

    The findings provide actionable insights for satellite developers and researchers, while highlighting the need for open-source whole-satellite platforms.

  • Takeaways & Limitations

    Manual taxonomy construction may introduce researcher subjectivity and affect validity despite independent classification and expert arbitration.

Abstract

from arXiv · show

Satellites have become fundamental components of modern technological systems, supporting critical infrastructure in communication, navigation, Earth observation, and scientific research. As space exploration advances and demand for satellite-enabled services grows, reliance on complex, heterogeneous satellite software continues to increase. A systematic understanding of the satellite software landscape is therefore increasingly important, yet existing studies still lack a comprehensive empirical examination. To address this gap, we present the first characterization study of open-source satellite software, examining its ecosystem and development practices. We mine and analyze 22,286 satellite-related GitHub projects through three research questions on popularity trends (RQ1), software goals (RQ2), and development practices (RQ3). First, we characterize the temporal evolution of projects and active developers, revealing increasing popularity. Second, through manual inspection of 646 projects, we construct a taxonomy of 43 software-goal categories spanning conceptual design, datasets, system implementation, simulation, testing, and tools. Third, we conduct an in-depth analysis of projects with source code, revealing a highly heterogeneous and task-specialized ecosystem with 66 programming languages and diverse implementation strategies. Finally, we summarize key findings and derive actionable implications for satellite developers and researchers.

1 Introduction

This study provides the first comprehensive empirical characterization of open-source satellite software by analyzing 22,286 GitHub projects. It examines popularity trends, software goals, and development practices, revealing sustained growth, broad objectives, and highly diverse implementation approaches.

  • Study Scope: 22,286 OSS satellite projects hosted on GitHub are analyzed to address research questions on popularity trends, goal taxonomy, and development practices.The study uses GitHub as its empirical basis because of the platform’s diversity and representativeness.
  • RQ1: Popularity Trend: Sustained and significant growth in community attention indicates the increasing prominence of satellite-related software topics.This finding underscores the timeliness and necessity of systematically studying OSS satellite projects.
  • RQ2: Goal Taxonomy: 646 manually inspected projects yield a taxonomy of 43 goal categories spanning datasets, system implementation, simulation, testing, and diverse tools.The taxonomy captures the broad range of functional objectives represented in OSS satellite projects.
  • RQ3: Development Practices: 66 distinct programming languages are used, with language selection closely aligned with task requirements and implementation strategies increasingly integrating heterogeneous models and solutions.The source-code analysis characterizes development practices across the satellite software ecosystem.
  • Implications: The study derives key findings and actionable implications for satellite software developers and researchers from its analyses.The work addresses the gap left by prior studies that largely overlooked a systematic understanding of the satellite software landscape.

2 Background and Related Work

Satellite software differs from traditional terrestrial software because distinctive environmental, operational, and mission constraints shape its design, implementation, and execution. Although prior research has examined satellite networks, devices, communication, security, and data labeling, open-source satellite software remains underexplored.

  • Satellite Software vs. Traditional Software: Satellite software operates under distinctive environmental, operational, and mission constraints that shape its design, implementation, and execution.Limited power, thermal capacity, and onboard hardware impose strict bounds on computation and storage.
  • Related Studies on Satellites: Prior satellite studies examine networks, devices, communication, security, and data labeling, but the broader software landscape remains insufficiently understood.The paper motivates systematic characterization to clarify the current state and provide actionable insights.
  • Related Studies on Open-Source Software: OSS research commonly uses GitHub repositories to study diverse domains, while open-source satellite software remains underexplored.Existing OSS studies span AIOps, data science, Blockchain, developer social networks, machine learning, and mobile app economy.

3 Methodology

The study follows prior empirical work to characterize open-source satellite projects through three research questions on popularity trends, software goals, and development practices. It analyzes 22,286 GitHub repositories, samples 646 projects for goal taxonomy construction, and examines source-code-available projects for language and implementation practices.

  • Research questions: The methodology addresses popularity trends, software-goal taxonomy, and development practices through RQ1, RQ2, and RQ3.The research questions respectively examine popularity trends, the specific goals addressed by projects, and adopted development practices.
  • RQ1: Popularity Trend: 22,286 repositories form the dataset for analyzing annual project creation and active developers in RQ1.New projects are identified by created_at timestamps, while annual active developers are counted as unique commit authors per year across projects.
  • RQ2: Goal Taxonomy: 646 projects are sampled for RQ2 with a 99% confidence level and ±5% margin of error.The large repository population makes exhaustive labeling impractical, so the study uses a statistically representative random subset.
  • RQ2: Goal Taxonomy: 453 projects support inductive taxonomy development, while 193 remaining projects are used to refine and extend the categories.Two authors collaboratively inspect descriptions, README files, source code, and related files, then independently label the remaining projects; Cohen’s Kappa is 0.857.
  • RQ2: Goal Taxonomy: 620 samples comprise the final taxonomy after 34 false labels, 604 single-category assignments, and 8 dual-category assignments.Popularity analysis of the 620 samples shows similarly increasing trends in annual project creation and active developers, suggesting robustness to sampling and false positives.
  • RQ3: Development Practices: RQ3 analyzes source-code-available projects across the 620 samples, focusing on programming-language and implementation practices.Language usage is derived from per-project language statistics, normalized across multi-language projects, and aggregated by goal category.

4 RQ1: Popularity Trend

Open-source satellite software has gained increasing attention, with growth in newly created projects, active developers, and development activity through 2025. These trends underscore the timeliness of the study and the need for software engineering support in this domain.

  • Project and developer growth: 4,997 newly created satellite projects were recorded in 2025, a 42.24% increase compared to 2024.The annual count rose from 1 project in 2008.
  • Project and developer growth: 9,018 annual active developers participated in satellite software projects in 2025, up from 1 in 2008.This growth accompanies the increase in newly created projects.
  • Development activity: 37,246 cumulative active developers and 136,248 annual commits were recorded by the end of 2025.Annual commits included 97,893 in 2020, a 35.89% increase over 2019.
  • Implications: Increasing attention and development activity underscore the timeliness of this study and the need for software engineering support in satellite software.The findings characterize the ecosystem as increasingly open and active.

5 RQ2: Goal Taxonomy · 5.1 Conceptual Design (A) · 5.2 Dataset (B)

The taxonomy reveals a highly diverse open-source satellite software ecosystem, spanning 22 inner categories and 43 leaf categories. This section covers conceptual-design projects and dataset resources, including observation data as the largest dataset subcategory.

  • 5 RQ2: Goal Taxonomy: 22 inner categories and 43 leaf categories organize OSS satellite projects across conceptual design, datasets, system implementation, simulation, testing, and tools.Leaf nodes represent the most fine-grained categories, while inner categories group multiple subcategories.
  • 5.1 Conceptual Design (A): 0.48% of projects fall under Conceptual Design, which presents satellite-related concepts or designs without concrete code implementations.These projects mainly serve illustrative purposes; one example is a satellite design supported by multiple illustrative documents.
  • 5.2 Dataset (B): 3.39% of surveyed projects belong to Dataset, providing satellite-related datasets for developer and researcher analysis and experimentation.The datasets are classified into four primary types.
  • 5.2 Dataset (B): 1.61% of projects provide Observation Data, including remote-sensing images or videos for applications such as land-cover and river-inundation analysis.Examples include land-cover imagery, river inundation imagery, and infrared videos for small-target detection.
  • 5.2 Dataset (B): 0.32% of projects provide Positioning Data within the Dataset category.Positioning Data is identified as one of the four primary dataset types.
  • 5.2 Dataset (B): Observation Data forms the largest Dataset subcategory and provides satellite-captured remote-sensing datasets.Overall, 3.39% of OSS satellite projects provide dataset resources.

5.3 System Implementation (C) · 5.4 Simulation (D)

System Implementation comprises 14.03% of surveyed projects across mission- and hardware-related systems, while Simulation comprises 7.42% and uses five simulation types. The ecosystem spans high-level mission functions, hardware interfaces, and controlled modeling of satellite systems and environments.

  • 5.3.1 Mission-related Systems (C.1, 8.87%).: Satellite Control Systems account for 2.74%, including Satellite Tracking (2.10%) and Satellite Attitude Control (0.65%).These systems determine and monitor orbital positions and manage satellite position and orientation in space.
  • 5.3.1 Mission-related Systems (C.1, 8.87%).: 4.19% of projects are Satellite-Ground Systems, the largest portion of Mission-related Systems, facilitating data exchange and operational coordination.Mission-related types also include satellite control operations, onboard communication and network management, and satellite-ground connectivity.
  • 5.3.2 Hardware-related Systems (C.2, 5.16%).: 3.55% of projects are Control Systems, the largest Hardware-related Systems subcategory, comprising Satellite Bus Control (1.77%) and Communication Signal Control (1.77%).Hardware-related Systems also include Onboard Data Handling Systems (0.65%) and Integration Systems (0.97%).
  • 5.3 System Implementation (C): 14.03% of surveyed projects belong to System Implementation, divided into Mission-related Systems (8.87%) and Hardware-related Systems (5.16%).Mission-related Systems support high-level mission functions, whereas Hardware-related Systems provide low-level support and hardware interfaces.
  • 5.4 Simulation (D): 7.42% of projects belong to Simulation, which models and emulates satellite systems, processes, or environments under controlled conditions.The category indicates the use of experimental validation in satellite software design and evaluation.
  • 5.4 Simulation (D): Five simulation types are employed, with Orbit simulation the most common at 3.06% and Satellite Environment Simulation the least common at 0.32%.Environment simulation models physical conditions affecting satellite operations, including electric fields and thermal environments.

5.5 Testing (E)

Testing projects evaluate whether satellite-related software, algorithms, or operational functions meet expected requirements, accounting for 1.94% of surveyed projects. Task-oriented algorithm testing is most prevalent alongside system performance testing, while calibration and stability testing each represent 0.16%.

  • Testing (E): 1.94% of surveyed projects belong to Testing, which evaluates software behavior, performance, accuracy, or robustness against expected requirements.Testing differs from Simulation by assessing whether components, algorithms, or operational functions satisfy requirements rather than modeling systems or environments.
  • Testing (E): 0.81% of surveyed projects conduct Task-Oriented Algorithm Testing (E.2), assessing image-processing and orbit-determination algorithms.E.2 is among the most prevalent Testing subcategories.
  • Testing (E): 0.16% of surveyed projects conduct Calibration Testing (E.3), examining the calibration accuracy of onboard instruments.Calibration Testing is one of the less prevalent Testing subcategories.
  • Testing (E): 0.16% of surveyed projects conduct Stability Testing (E.4), focusing on the stability of flight software.Stability Testing is one of the less prevalent Testing subcategories.

5.6 Tool (F)

Tools form the largest satellite-software category, comprising 72.74% of projects and spanning data utilities, spatial and orbital analysis, communications, and visualization. Data Utility dominates the category at 53.71%, centered on satellite data collection and processing.

  • Data Utility: 41.77% of projects perform Imagery Processing, including scenario-specific analysis at 22.90% and scenario-independent processing at 18.87%.Scenario-specific imagery processing includes 142 projects, while scenario-independent processing includes 117 projects and covers preprocessing, enhancement, analysis, and visualization.
  • Spatial and Orbital Analysis: Spatial and orbital tools support coverage analysis at 0.65%, geolocation analysis at 3.06%, scheduling at 0.65%, orbit determination at 0.81%, and orbit propagation at 2.74%.These tools analyze observed or serviced areas, estimate observation positions, plan overpasses, estimate orbital elements, and predict future orbital states.
  • Communication and Visualization: Communication tools address link analysis at 0.65%, protocol analysis at 0.65%, and signal analysis at 1.61%, while visualization tools cover trajectories at 5.65% and satellite data at 2.58%.Communication tools assess links, protocols, and signal quality; visualization tools render orbital paths, temporal motion, and analytical outputs.
  • Tool: 72.74% of projects belong to the Tool category, the taxonomy’s largest category, spanning 9 inner and 16 leaf categories.The category comprises modular, function-oriented tools and includes five subcategories.
  • Tool: 53.71% of the Tool category is Data Utility, covering satellite data collection and processing for non-imagery and imagery data.Data Collection accounts for 4.68%, including Data Retrieval at 2.42% and Data Download at 2.26%; Data Processing accounts for 49.03%.

6 RQ3: Current Development Practices

Open-source satellite software follows a heterogeneous, task-aligned development landscape spanning 66 programming languages and diverse implementation strategies. Projects address mission operations, hardware integration, simulation, and performance and algorithm testing through both classical and data-driven methods.

  • Programming languages: 66 distinct languages reveal substantial technological diversity, with Python, Jupyter Notebook, JavaScript, C++, MATLAB, and HTML the six most prevalent.Their shares are 32.91%, 28.06%, 7.40%, 4.40%, 4.06%, and 3.57%, respectively.
  • Programming languages: Programming choices vary by task: system implementation combines Python with lower-level languages, whereas simulation favors Python and MATLAB and testing and tools favor interactive and scripting ecosystems.In System Implementation, Python accounts for 29.22%, followed by C++ at 14.95% and C at 9.38%; in Simulation, Python accounts for 42.07% and MATLAB 21.81%.
  • System implementation: Mission-related systems support orbit tracking, attitude control, signal processing, robust communication, adaptive routing, scheduling, power management, and satellite-data storage, analysis, and visualization.Implementations combine SGP4 and PD control with reinforcement learning, while operational systems also cover telemetry, imagery, and spatial-information workflows.
  • System implementation: Hardware-related systems span PCB- and microcontroller-based onboard data management, CubeSat integration, bus control, signal acquisition, RF integration, and processing pipelines.Examples include custom PCBs, dedicated onboard computers, Arduino, STM32 architectures, Raspberry Pi, commercial receivers, and GNU Radio.
  • Simulation: Simulation projects model environments, attitude, orbits, constellations, subsystem interactions, communication, networking, and mission planning using physical dynamics, control inputs, and collaborative-operation models.Methods include solar-radiation and material-property models, quaternion dynamics, SGP4 and Keplerian propagation, thruster inputs, TLE configurations, federated learning, protocol simulation, and link-budget estimation.
  • Testing: Testing evaluates network and communication-link performance alongside mission-specific image-processing and estimation algorithms under realistic channels, sensor noise, and satellite-like constraints.Targets include shoreline extraction, orbit determination, attitude estimation, and space situational awareness.

7 Discussion and Implications

The discussion identifies dominant observation-data projects, growing need for configurable satellite software under onboard constraints, and an immature ecosystem lacking open-source end-to-end platforms. It translates these findings into implications for satellite developers and researchers.

  • Satellite Datasets: Observation data dominate open-source satellite dataset projects, enabling large-scale imagery datasets and reflecting the mission-critical role of Earth observation satellites.
  • Actionable Insights: Strict computation, energy, and bandwidth constraints make efficient configuration essential, supporting configurable architectures and runtime-adjustable parameter management.
  • Research Implications: The open-source satellite software ecosystem remains immature, concentrated on specialized tools and partially open components, with end-to-end satellite platforms largely absent.
  • Research Implications: Open-source whole-satellite platforms could enable end-to-end experimentation, system-level evaluation, and cross-layer analysis of computation, communication, control, and environmental interactions.

8 Threats to Validity

Manual taxonomy construction may introduce researcher subjectivity, although independent classification, discussion, expert arbitration, and high inter-rater agreement were used to mitigate this threat.

  • Researcher Subjectivity: Manual analysis used to construct the satellite software-goal taxonomy may introduce researcher subjectivity and affect validity.
  • Researcher Subjectivity: Two authors independently classified projects, resolving disagreements through discussion and satellite-domain expert arbitration when necessary.
  • Researcher Subjectivity: 0.857 Cohen’s Kappa indicates a high level of agreement between the independent classifications.

9 Conclusion

This paper provides the first empirical characterization of open-source satellite software, analyzing 22,286 GitHub repositories to examine popularity trends, software goals, and development practices. It finds sustained growth in projects and active developers and derives 43 software-goal categories from 646 sampled projects.

  • 22,286 satellite-related GitHub repositories were analyzed across popularity trends, functional objectives, and development practices.
  • 646 projects were manually sampled to derive a taxonomy of 43 software goals spanning datasets and system implementation.
  • Sustained growth in satellite software and active developers underscores the rising importance of open-source ecosystems in this domain.
Loading 2608.26211v1…