Identity authentication based on trajectory characteristics of mobile devices

Abstract

Users’ trip location information is of great value in the study of social computing and big data. Based on the personalized characteristics of user trajectory data, this paper proposes a Trajectory-based Identity Authentication Method (TIAM). TIAM consists of the registration phase and the authentication phase. Users initialize the trip routes and the coordinates of home and office in the registration phase to set template trajectories and stay points. In the authentication phase, TIAM first preprocesses trajectory data, and then calculates the similarity between the sample trajectories to be authenticated and the template trajectories in the template library, and detects if there are stay-point hits in the sample trajectories. Third, TIAM authenticates whether the current user is valid based on similarities and stay-point hits, and updates the template library with new sample trajectories that are authenticated as valid to accommodate changes in the users’ trip routes. Because the process of collecting data and authentication is transparent to a user, TIAM does not reveal the user’s biological information, and can dynamically adapt to the change of user’s trajectories. Therefore, TIAM has strong flexibility and adaptability. Experiments show that the average authentication accuracy of TIAM is 96.74%. It is not only suitable for the identification of mobile device owners, but also suitable for monitoring children safety and transport vehicles of dangerous goods.