[Colloq] Master's thesis defense - Pan Qiu - Malicious Traffic Analysis of Android Applications - Wednesday, April 29- 1 pm - 166 WVH

[DiFazio, Danielle]

Who: Tongzu Ding

What: Master's Thesis Defense

When: Wednesday, April 29, 2015, 1 PM

Where: 166 WVH

Title: Malicious Traffic Analysis of Android Applications

Abstract:

We have witnessed the explosive growth of smartphone adoption in recent years, with the Android operating system becoming one the most popular platforms. While this platform makes it easy for app developers to develop and release legitimate apps, it is also easy for malicious parties to use Android to spread malware. In response, researchers have developed numerous static and dynamic analysis techniques to mitigate this issue. However, most existing techniques concentrate on malware app activities on a single host and fail to analyze their network behavior.
 
In this paper, we focus on analyzing network traffic of Android malware. Specifically, we tested more than 1400 Android malware samples, collect their runtime network traffic, summarize the characteristics and extract network features for classification. Based on the evaluation with different classifiers, our experiments shows an overall 73% accuracy as well as 95.0%, 90.7% and 86.8% accuracy for dowgin, provar and smsreg malware family respectively. We also find that classification accuracy is related to traffic generation rate of malware family.
Committee:

Thesis Committee: 
David Choffnes
Themis Papageorge