2021 · Conference paper
Asteria: Deep Learning-based AST-Encoding for Cross-platform Binary Code Similarity Detection
ASTERIA uses a Tree-LSTM to encode function ASTs and compare binary functions across platforms. It was evaluated on function pairs and used to search IoT firmware for vulnerabilities.
Scope and limitations
The model was trained on 1,022,616 function pairs and evaluated on 95,078 pairs. Its cross-platform scope covers differing IoT architectures; a firmware search identified 75 vulnerable functions in 5,979 images.
Why it fits this collection
It studies reverse engineering through binary function similarity and vulnerability search in IoT firmware.
Publication and review details
- Venue
- 2021 51st Annual IEEE/IFIP International Conference on Dependable Systems and Networks (DSN)
- Status
- Conference paper
- Publication / posting date
- June 2021
- Date source
- Published date in the Crossref record for the DOI listed on arXiv.
- First arXiv submission
- August 13, 2021
- arXiv
- 2108.06082
- Reviewed arXiv version
- v1
- DOI
- 10.1109/dsn48987.2021.00036
The DOI comes from the arXiv record, and the venue and date from Crossref.
Sources reviewed
Reviewed against the paper and its primary sources. Cite the original authors using the linked paper or BibTeX.