Extensible and Scalable Architecture for Hybrid Analysis

The prevalence of Android apps and their widespread use in daily life has made them a prime subject of study in program analysis. Apps for e-mail, navigation, mobile banking, eGovernment, healthcare, etc. each have their respective requirements on stability, effciency, and security, which can be checked using static and dynamic analysis. While developers and researchers can pick from a variety of scalable and integrated frameworks for static analysis, designing a dynamic analysis still requires significant engineering and design effort that contributes little to the analysis task at hand. Existing scholarly work on dynamic analysis has instead focused on individual challenges such as effcient data flow tracking, or code coverage in UI exploration. Combining dynamic analysis configuration and results with artifacts from static analysis is usually dealt with on an individual basis that does not generalize in the sense of a re-usable framework. In this paper, we present a reference architecture and implementation for an integrated, scalable, and extensible hybrid analysis that offers a wide range of dynamic analysis capabilities. We hope that researchers can build upon our work for increased effciency in hybrid analysis.