Classification and Feature Extraction for User Identification for Smart Home Networks Based on Apps Access History

Advancements in smartphones has led to increasing dependency on them as an essential part of IoT smart home networks. Although this increase provides convenience to users, many challenges, including user identification and authentication, need to be considered. Many related works consider smartphone user authentication that assumes that the same user will be using the device throughout the access session. However, after the login stage, the device could be used by others, either in the home environment or outside, leading to undesired access to home appliances. In this paper, we present a continuous user identification approach that uses implicit features that can be integrated as a second layer of identification beyond the login step, based on user behavior on smartphones. The proposed method is mainly based on user interaction on the smartphone, which thus reflects the user's pattern that in turn provides the impetus to employ user identification as a complementary approach beyond the user login stage.