A hybrid authentication system for websites on mobile browsers

Current biometric recognition systems for website authentication are mostly web server based, needing server support and infrastructure, and sometimes dedicated external hardware, for online authentication. Not many web servers support this kind of infrastructure for authentication. On the other hand, pure device based authentication systems are of the `all or none' type, serving to authenticate users for every action when using the device, or having authentication once and then keeping the device free to use. In this paper we propose a hybrid model, where fingerprint authentication is used in combination with the auto complete function on the browser for logging in to certain types of websites, or accessing only certain kind of information on the browser. In this method, the fingerprint module, inbuilt on certain mobile devices like Samsung Galaxy, is triggered automatically when certain pre-configured rules are met, such as the user visiting certain categories of websites. Also, based on the identity of the person swiping their fingerprint, the device has the ability to switch between one of a number of preconfigured security modes. Such a model can enable enhanced authentication for logging in to websites on mobile devices. We present the results of tests on a browser enabled with this system to study average response times, accuracy and effect on browser performance.