Space RadSim: Simulating Cosmic Radiation Effects on Hardened Satellite Firmware
Space RadSim is a deterministic, binary-agnostic fault injection tool that simulates cosmic radiation effects (bitflips) on satellite firmware to evaluate the tradeoffs between radiation resilience and exploit mitigation techniques.
Published at IEEE Symposium on Security and Privacy (S&P) 2025
Space RadSim is a deterministic, binary-agnostic fault injection tool that simulates cosmic radiation effects on satellite firmware equipped with exploit mitigation techniques. It evaluates the tradeoffs between radiation resilience and security defenses in space systems by injecting faults into satellite binaries compiled with various hardening flavors.
Key Features
- Fuzzing & Corpus Selection — Runs parallel fuzzing instances to generate test inputs, with configurable duration and core counts
- Coverage Analysis — Computes basic block coverage metrics post-fuzzing to validate test comprehensiveness
- Input Aggregation — Consolidates results from independent fuzzing runs and selects top-performing inputs by combined coverage
- Trace Generation — Creates single-trace executions for bitflip injection and cross-flavor comparisons across different mitigation strategies
- Fault Injection — Exhaustively injects single-bit flips into RAM regions to assess system resilience
How It Works
Users prepare target binaries organized by satellite name and mitigation flavor
(e.g., plain, canary-strong, cfi), then invoke run-radsim.py with appropriate
subcommands to execute fuzzing campaigns, analyze coverage, and perform fault
injection experiments.
Built primarily in Rust with Python scripting.
