Characterisation of unsolicited traffic advertisements in mobile devices

Abstract

Smartphones play an undeniable and increasingly important role in people's life, whether to communicate or perform tasks requiring more computing capacity. This growth is also associated with an increase in network traffic due to the widespread use of mobile applications. To have global Internet access, users need to contract a mobile data plan to complement Wi-Fi connectivity, normally, limited to a certain traffic volume. Once exhausted either the data plan is extended, sometimes with a prohibited cost, or the data rate is reduced significantly being often inadequate for many applications. In particular, the veiled presence of unsolicited traffic contributes for mobile data plan consumption. This justifies the need for studying and understanding unwanted traffic exchanged between the mobile device and the network in order to assess how this traffic type impacts on the user data plan. To answer this need, a systematic and multilevel methodology is developed and proposed for characterising unsolicited traffic of type advertisements in mobile devices. The study considers as inputs popular user applications - YouTube, Facebook and Instagram - which may have substantive impact on the user data plan. This work also sheds light on how ad serving is accomplished in these applications.