Behavioral biometrics is a scientific approach to user authentication/identification based on how a user interacts with the authenticating device. Typical instances of behavioral biometrics include signature, voice, and keystroke dynamics. Each of these approaches provides a moderate to high level of security – depending on the particular instantiation. More recent advances in this technology include the deployment of biosignals: electrocardiogram and electroencephalogram data as means of authentication. Error rates with these technologies are comparable to more traditional behavioral biometrics and indeed even moderate physiological based biometrics such as fingerprint scanning. By combining a variety of behavioral biometrics (e.g. keystroke dynamics and EEG) – a multi-modal approach – error rates can be reduced further still. Recent technology has provided a complete platform based on a virtual reality environment that provides unique challenge based authentication approaches that provide an extremely robust approach with optimal error rates. These issues will be discussed through a case-based approach. |