Diagnosis of Privacy and Performance Problems in the Context of Mobile Applications.

Complex systems need to run flawless and protecting the users privacy is mandatory to stay competitive. The consideration of privacy aspects is even more important for mobile applications, because a lot of sensitive data is transferred over a bunch of different networks. Especially after the espionage scandals in the last years the users care about their privacy more than ever. There are several approaches which enable the monitoring and analysis of performance issues. However, there is no approach which supports both privacy and performance diagnostics. Therefore we propose an approach for filling this gap, a monitoring and analysis strategy for applications which is capable of identifying privacy and performance problems. The monitoring is designed to collect performance metrics together with privacy related information. The resulting monitoring data is used to derive an architectural run-time model which takes part in the performance analysis and in the detection of privacy threats. During our evaluation we analyzed the accuracy and the overhead to confirm that our approach can be used in production.