Behavioral analysis for a continuous user authentication

New generation devices are pervasive in nature and provide a number of security sensitive functionalities, which might expose the user private information to serious security and privacy threats. The main countermeasure used to prevent unauthorized access is the user authentication. Most of these devices are still protected by traditional authentication mechanisms (PIN, password), which are exposed to well known security limitations. These issues are mitigated by the introduction of new physical biometric authentication mechanisms. Biometric authentication, basing on user physical traits and requiring the user presence at the authentication time, makes the system more secure. Despite the new mechanisms overcome some data security issues, they still suffer from other usability problems. In this paper we explore a new unobtrusive authentication mechanism based on human behavior.