AndroGenerator: An automated and configurable android app network traffic generation system. Issue 18 (15th September 2015)

• Record Type: Journal Article
• Authors: Su, Xin; Zhang, Dafang; Li, Wenjia; Wang, Xiaofei
• Is Part Of: Security and communication networks. Volume 8:Issue 18(2015)
• Page Start: 4273
• Page End: 4288
• Publication Date: 2015-09-15
• Subjects: Android application -- traffic generator -- automatic execution; Computer networks -- Security measures -- Periodicals; Computer security -- Periodicals; Cryptography -- Periodicals; 005.805
• DOI: 10.1002/sec.1341
• Languages: English
• ISSNs: 1939-0114

Abstract

Abstract: With the rapid growth in the popularity of Android smartphones, a large number of Android applications (or apps) have emerged in both official and alternative Android markets. It is important for network operators and security analysts to understand the network traffic generated by new Android apps for the purposes of network management, app traffic analysis, and malware detection. However, it is time‐consuming and tedious to manually install and run Android apps to generate network traffic. Moreover, existing synthetic network traffic generators are unable to generate network traffic that can accurately reflect the network behaviors of Android apps. In this paper, we propose and implement AndroGenerator, an automated Android network traffic generation system, to generate various types of network traffic that can be produced by Android apps. Our system reproduces the network traffic based on the traffic characteristics extracted from traffic traces captured by running a large number of Android applications from several popular Android markets, such as Google Play. The system first generates network traffic through automated execution of Android applications. Then, the system is also able to extract network characteristics from the captured traffic traces and store the extracted results into a database for benchmarking purposes. Finally, AndroGenerator reproduces Android app traffic based via simulating network characteristics of captured traffic traces. In the experiments, we evaluate our system with real‐world mobile traffic, and the experiment results show that AndroGenerator can reproduce Android app network traffic accurately. Copyright © 2015 John Wiley & Sons, Ltd. Abstract : In this paper, we propose and implement AndroGenerator, an automated Android network traffic generation system, to generate various types of network traffic that can be produced by Android apps. Our system reproduces the network traffic based on the traffic characteristics extracted from traffic traces captured by running a large number of Android applications from several popular Android markets. The experiment results show that AndroGenerator can reproduce Android app network traffic accurately.