Beyond Fuzzy Matching: Constraint-Guided Patch Presence Testing for Obfuscated Java Binaries
Lige Zhan, Jiang Ming, Chenke Luo, Letian Sha, Guojun Peng, Jianming Fu
Abstract
Patch presence testing (PPT) is critical for mitigating vulnerabilities introduced by the use of third-party libraries (TPLs) in Java programs. However, the widespread use of obfuscation significantly undermines its accuracy by altering patch-relevant features. To preserve feature consistency before and after obfuscation for reliable comparison, existing studies summarize patch digests using several representative fuzzy features. These fuzzy features, designed to generalize obfuscation-prone details, have proven effective in TPL detection. Nevertheless, this approach faces limitations when directly applied to PPT, where finer features are needed to capture subtle patch changes. Although patch digests facilitate feature matching under obfuscation-induced code changes with high efficiency, methods with similar semantic structures—common in TPLs due to factors such as overriding and inheritance—often exhibit identical digests. Relying solely on feature similarity without considering method constraints often results in inaccurate outcomes.
Ask about this paper
Ask your agent about it.
Lune has read the top-tier papers around this one, so every answer names the papers it rests on.
Related papers
- Precise and Efficient Patch Presence Test for Android Applications against Code ObfuscationZifan Xie, Ming Wen, Haoxiang Jia, Xiaochen Guo et al.ISSTA 2023 · 12 citations
- Towards Global Matches for Third-Party Library Detection in AndroidLige Zhan, Jiang Ming, Chenke Luo, Guojun Peng et al.ICSE 2026
- PPT4J: Patch Presence Test for Java BinariesZhiyuan Pan, Xing Hu, Xin Xia, Xian Zhan et al.ICSE 2024 · 6 citations
- LibScan: Towards More Precise Third-Party Library Identification for Android ApplicationsYafei Wu, Cong Sun, Dongrui Zeng, Gang Tan et al.USENIX Security 2023
- An Empirical Study on the Robustness of Android Third-Party Library Detection Tools Against Advanced ObfuscationDahan Pan, Zhuohao Zhang, Yunjia Min, Runhan Feng et al.ICSE 2026
