A Location-Aware Authentication Model to Handle Uncertainty in IoT

The process of authentication for location-aware services is more challenging over the Internet of Things (IoT) than before due to a number of inherent IoT characteristics like dynamism and heterogeneity. Authentication in such an environment is no longer dependent only upon credentials but rather needs more spatio-temporal parameters (i.e. Time, Location) to be considered. Moreover, big data produced by IoT enabling technologies like wireless sensors, RFID tags and mobile phones introduce new challenges in terms of real-time authentication. This work defines one of the neglected challenges in IoT called uncertainty in authentication. Considering this challenge is important because IoT has a number of characteristics that amplify uncertainty in authentication. Moreover, using existing access control models does not result in flexible and resilient solution to handle uncertainty. Therefore, this work introduces an uncertainty-aware authentication model based on Attribute-Based Access Control (ABAC), which uses spatio-temporal parameters as attributes. Our model can achieve authentication performance with 86.92% accuracy with lower overhead using Neural Networks.