[Paper Review] Open Source Android Vulnerability Detection Tools: A Survey

Keyur Kulkarni, Ahmad Y. Javaid|arXiv (Cornell University)|Jul 31, 2018

Advanced Malware Detection Techniques12 references3 citations

TL;DR

This survey analyzes open-source static and dynamic Android vulnerability detection tools, evaluating their capabilities, limitations, and design principles. It systematically categorizes tools, compares three leading ones—Androvigilante, DroidSafe, and MobSF—highlighting their strengths in identifying common Android security flaws such as insecure data storage and improper permissions.

ABSTRACT

Since last decade, smartphones have become an integral part of everyone's life. Having the ability to handle many useful and attractive applications, smartphones sport flawless functionality and small sizes leading to their exponential growth. Additionally, due to the huge user base and a wide range of functionalities, these mobile platforms have become a popular source of information to the public through several Apps provided by the DHS Citizen Application Directory. Such wide audience to this platform is also making it a huge target for cyber- attacks. While Android, the most popular open source mobile platform, has its base set of permissions to protect the device and resources, it does not provide a security framework to defend against any attack. This paper surveys threat, vulnerability and security analysis tools, which are open source in nature, for the Android platform and systemizes the knowledge of Android security mechanisms. Additionally, a comparison of three popular tools is presented.

Motivation & Objective

• To identify and categorize open-source tools designed for detecting vulnerabilities in Android applications.
• To analyze the security mechanisms and detection techniques employed by these tools to assess their effectiveness.
• To compare the performance, coverage, and usability of three prominent open-source Android vulnerability detection tools.
• To provide a systematic overview of the current state of the art in open-source Android vulnerability detection for researchers and practitioners.
• To highlight gaps in existing tools and suggest directions for future research in Android security analysis.

Proposed method

• Systematic literature review and analysis of open-source Android vulnerability detection tools from public repositories and academic sources.
• Categorization of tools based on their analysis approach: static, dynamic, or hybrid (static + dynamic) analysis.
• Evaluation of tools based on supported vulnerability types, accuracy, scalability, and integration capabilities.
• Comparison of three representative tools—Androvigilante, DroidSafe, and MobSF—using criteria such as detection coverage, performance, and ease of use.
• Mapping of detected vulnerabilities to common Android security issues, including insecure data storage, improper permissions, and insecure communication.
• Synthesis of findings into a structured framework for understanding tool capabilities and limitations in real-world Android security assessment.

Experimental results

Research questions

• RQ1 What are the key categories and characteristics of open-source Android vulnerability detection tools?
• RQ2 How do static and dynamic analysis techniques compare in detecting common Android security vulnerabilities?
• RQ3 What are the strengths and limitations of Androvigilante, DroidSafe, and MobSF in identifying security flaws in Android applications?
• RQ4 To what extent do existing tools cover known Android security vulnerabilities such as insecure data storage and improper permissions?
• RQ5 What gaps remain in current open-source tooling for Android vulnerability detection, and how can they be addressed?

Key findings

• The survey identifies and classifies multiple open-source tools that employ static, dynamic, or hybrid analysis to detect Android vulnerabilities.
• Androvigilante excels in detecting insecure data storage and improper permissions through static analysis of bytecode.
• DroidSafe provides strong taint analysis capabilities and is effective in identifying data leakage vulnerabilities via static program analysis.
• MobSF offers a comprehensive, web-based interface supporting both static and dynamic analysis, making it accessible for non-experts.
• No single tool covers all vulnerability types with high accuracy, indicating a need for complementary tool usage.
• The study reveals that while open-source tools have made significant progress, challenges remain in scalability, precision, and coverage of complex attack patterns.