Design Of Authentication System Based On Device Identification

With the developing of network and information technology, it alsobrings security problem to us. Legacy authentication system can’t protect usfrom network fraud attack. In order to harden authentication system security,we need to introduce multi-factor authentication. As one kind ofauthentication methods, recently device identification has proven useful in thedetection and prevention of online identity theft and credit card fraud.This paper takes advantage of passive device identification, based onmixed configuration information. Because the configurations are changingover time in a network, we need to calculate the device credibility when userauthenticating. This paper introduce Bayesian method to achieve this task,comparing with the static similarity pattern match in previous research, thismethod provides a more accurate result. In combine with this deviceidentification technique, an authentication system is designed. For the sake oftesting logistic functions, we also built a simulation system, and tested it withfour cases. The test result was successful.This paper integrated device identification technique withauthentication system, this enables the authentication system performingmulti-factor authentication while not affecting user experience. We alsointroduced Bayesian method to calculate device’s credibility, this workprovides a more accurate credibility result. Because the calculation is basedon statistics data collected, it is a self-learning system.