Computer Science and Engineering, Department of


Date of this Version



Published in Journal of Computer Languages 52 (2019), pp 55–71.

doi 10.1016/j.cola.2018.12.004


Copyright © 2019 Elsevier Ltd. Used by permission.


Security vetting of Android apps is often performed under tight time constraints (e.g., a few minutes). As such, vetting activities must be performed “at speed”, when an app is submitted for distribution or a device is analyzed for malware. Existing static and dynamic program analysis approaches are not feasible for use in security analysis tools because they require a much longer time to operate than security analysts can afford. There are two factors that limit the performance and efficiency of current analysis approaches. First, existing approaches analyze only one app at a time. Finding security vulnerabilities in collaborative environments such as Android, however, requires collaborating apps to be analyzed simultaneously. Thus, existing approaches are not adequate when applied in this context. Second, existing static program analysis approaches tend to operate in a “closed world” fashion; therefore, they are not easily integrated with dynamic analysis processes to efficiently produce hybrid analysis results within a given time constraint.

In this work, we introduce JITANA, an efficient and scalable hybrid program analysis framework for Android. JITANA has been designed from the ground up to be used as a building block to construct efficient and scalable program analysis techniques. JITANA also operates in an open world fashion, so malicious code detected as part of dynamic analysis can be quickly analyzed and the analysis results can be seamlessly integrated with the original static analysis results. To illustrate JITANA’s capability, we used it to analyze a large collection of apps simultaneously to identify potential collaborations among apps. We have also constructed several analysis techniques on top of JITANA and we use these to perform security vetting under four realistic scenarios. The results indicate that JITANA is scalable and robust; it can effectively and efficiently analyze complex apps including Facebook, Pokémon Go, and Pandora that the state-of-the-art approach cannot handle. In addition, we constructed a visualization engine as a plugin for JITANA to provide real-time feedback on code coverage to help analysts assess their vetting efforts. Such feedback can lead analysts to hard to reach code segments that may need further analysis. Finally we illustrate the effectiveness of JITANA in detecting and analyzing dynamically loaded code.

Supplementary material attached below.