Adaptive multi-factor authentication system
A system and methodology for adaptive selection of multiple modalities for authentication in different operating environments, thereby making authentication strategy unpredictable so to significantly reduce the risk of exploitation by authentication-guessing attacks. The system calculates trustworthiness values of different authentication factors under various environmental settings, and combines a trust-based adaptive, robust and scalable software-hardware framework for the selection of authentication factors for continuous and triggered authentication with optimal algorithms to determine the security parameters of each of the authentication factors. A subset of authentication factors thus are determined for application at triggering events on-the-fly, thereby leaving no exploitable a priori pattern or clue for hackers to exploit.
1. A machine for improved secure access to computing devices, systems, resources, or services, comprising:
one or more computer servers with authentication modality data stored thereon for a plurality of authentication modalities, wherein the authentication modality data for each authentication modality comprises a trustworthiness factor for each of one or more user input devices, a trustworthiness factor for each of one or more user connection media, and a computational complexity cost factor; and
a processor or microprocessor, wherein the processor or microprocessor is programmed to determine one or more of said authentication modalities to use for an authentication verification event by:
determining the objective trustworthiness value for each modality based on device trustworthiness factors and connection media trustworthiness factors for said modality;
determining a penalty value for each modality based on the computation complexity cost factor for said modality and the previous selection history of said modality for previous authentication verification events;
ranking the authentication modalities based on the objective trustworthiness value and the penalty value; and
applying one or more authentication modalities in order of ranking.
2. The machine of claim 1 , wherein the user input devices comprise fixed devices, mobile devices, hand-held devices, or combinations thereof.
3. The machine of claim 1 , wherein the user media environments comprise wired, wireless, cellular, or combinations thereof.
4. The machine of claim 1 , wherein authentication modality data for each authentication modality comprises a list of features applicable to that authentication modality.
5. The machine of claim 4 , wherein one or more features from said list of features are selected for an authentication modality for determining the objective trustworthiness value for that modality.
6. The machine of claim 1 , wherein said authentication modalities comprise one or more of the following modalities: facial recognition, fingerprint recognition, password, CAPTCHA, voice recognition, and keystroke analysis.
7. The machine of claim 1 , wherein the plurality of authentication modalities comprises seven modalities: facial recognition, fingerprint recognition, password, CAPTCHA, voice recognition, and keystroke analysis.
8. The machine of claim 1 , wherein the processor or microprocessor further is programmed to modify one or more trustworthiness factors.
9. The machine of claim 1 , wherein the authentication verification event is an initial request by a user using a user input device to access a computer device, system, resource, or service.
10. The machine of claim 1 , wherein the authentication verification event is an automatic determination of whether to allow a user using a user input device to continue having access a computer device, system, resource, or service.