Network analytics of mobile apps
Smartphones rely on encryption to maintain the security and privacy of mobile apps, which has made network management a complex task.
Due to the proliferation of smartphones, the need to manage mobile devices and their applications on an enterprise network has never been so critical.
The goals for the project were to analyse the behaviour of mobile devices and their applications to determine whether they can be classified using machine learning, a form of artificial intelligence.
To achieve this, user actions on various mobile apps were automated on a testbed of 7 devices to collect labelled network traffic data. Machine learning algorithms were then utilised to classify a closed set of mobile apps and operating systems. Our findings were that Android apps along with iOS and Android operating systems can be classified to a high degree of accuracy.
Theme
Securing our future
Booth
SF47
School
Electrical and Electronic Engineering
Exhibitors
Leon Dimitrakopoulos
Leo Hague
Dave Rattan