The purpose of this paper is to extract the user behaviour and transform it into a unique signature that can be used as implicit authentication technique. Smart devices are equipped with multiple authentication techniques and still remain prone to attacks because all of these techniques require explicit intervention of the user. Entering a pin code, a password or even having a biometric print can be easily hacked by an adversary.
In this paper, the authors introduce a novel authentication model to be used as complementary to the existing authentication models. Particularly, the duration of usage of each application and the occurrence time were examined and modelled into a user signature. During the learning phase, a cubic spline function is used to extract the user signature based on his/her behavioural pattern.
Preliminary field experiments show a 70 per cent accuracy rate in determining the rightful owner of the device.
The main contribution of this paper is a framework that extracts the user behaviour and transforms it into a unique signature that can be used to implicitly authenticate the user.
