Home // SECURWARE 2018, The Twelfth International Conference on Emerging Security Information, Systems and Technologies // View article

Exploiting User Privacy in IoT Devices Using Deep Learning and its Mitigation

Authors:
Rana AlAmeedee
Wonjun Lee

Keywords: Internet of Thing; Smart Home; Privacy; Deep Learning

Abstract:
Internet of Things (IoT) has seen a great growth in recent years; the number of devices is expected to be 80 billion by 2025. Although the IoT facilitates our life, however, it threatens our privacy if we do not take the necessary security measures. In this paper, we show how the user activities can be tracked using only network traffic packets sent from several commercial IoT devices with no need for deep inspection. The prediction about daily life activities of the user at home is made based on analysis of deep learning. In addition, we propose a practical idea to mitigate the privacy attack caused by the smart home devices, and introduce experimental results showing that our approach works very accurately.

Pages: 43 to 47

Copyright: Copyright (c) IARIA, 2018

Publication date: September 16, 2018

Published in: conference

ISSN: 2162-2116

ISBN: 978-1-61208-661-3

Location: Venice, Italy

Dates: from September 16, 2018 to September 20, 2018