Boosting Grey-box Fuzzing for Connected Autonomous Vehicle Systems
Bibliographic record
Abstract
Assuring the cybersecurity of Connected Autonomous Vehicles (CAVs) entails protecting the data, devices, network connectivity, and, most importantly, autonomous vehicles' software. Software security testing aims to minimize the attack surface of CAVs by identifying security vulnerabilities at an early stage. One of the most robust and efficient security testing methods is fuzzing. Though fuzz testing can validate the system with various scenarios, its blindness prevents it from exploring the deep paths of the system. Hence, the automotive industry needs a reliable security testing tool that dynamically explores the system and assures a comprehensive evaluation. This paper presents a hybrid fuzz testing framework (VulFuzz++) that unites the efficiency of fuzzing and the precision of concolic execution to provide the automotive industry a reliable security testing tool. VulFuzz++offloads most of the exploration process to the vulnerability-oriented fuzzer (VulFuzz) explicitly designed for automotive systems. When the fuzzer halts failing to explore different paths, VulFuzz++examines the untraversed branches and prioritizes them based on their potential to expose vulnerabilities. It utilizes a tailored, targeted concolic engine that limits the symbolic exploration to only specific functions. When the concolic engine discovers new system inputs, testing is handed over again to the fuzzer to perform a quick and efficient evaluation of the newly explored region. We implemented and experimented with the VulfFuzz++framework on a driving assistance system. VulFuzz++boosted the vulnerability exposure process of grey-box fuzzing, increasing the obtained crashes by 50%. It dramatically outperforms traditional concolic engines in assisting fuzzers, exposing 50 times more unique crashes. VulFuzz++extends the testing time moderately but assures a comprehensive examination covering 96.7% of the automotive system branches.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".