Secure authentication using recurrent neural networks

1. A computer-implemented method of user authentication, the method comprising:

using a number of processors to perform the steps of:

combining, by a computer system, a user recurrent neural network with a system recurrent neural network to form a unique combined recurrent neural network, wherein the user recurrent neural network is configured to generate a unique user key and the system recurrent neural network is configured to generate a system key;

inputting, by the computer system, a predetermined input into the combined recurrent neural network;

generating, by the combined recurrent neural network, a unique combined key from the input, wherein the combined key differs from both the user key and system key; and

associating, by the computer system, the combined key with a unique access authorization to authenticate a user.

2. The method of claim 1 , further comprising inputting a second predetermined input to the combined recurrent neural network for multi-factor authorization.

3. The method of claim 2 , wherein the second predetermined input is provided by the user.

4. The method of claim 3 , wherein the second predetermined input comprises one of:

a password;

a passphrase;

personal identification number;

voice recognition;

facial recognition;

retinal scan;

iris recognition;

fingerprints;

handprint; or

palm vein pattern.

5. The method of claim 2 , wherein the second predetermined input is provided by the computer system.

6. The method of claim 1 , wherein the user recurrent neural network and system recurrent neural network comprise liquid state machines.

7. The method of claim 1 , wherein the user recurrent neural network and system recurrent neural network comprise neuromorphic hardware.

8. The method of claim 1 , wherein the predetermined input comprises a pre-determined sequence of neuron spikes.

9. The method of claim 1 , wherein the combined recurrent neural network comprises a physically unclonable function.

10. The method of claim 1 , wherein the user recurrent neural network and system recurrent neural network combined produce encrypted output data as a function of input data and the structures of the recurrent neural networks.

11. A computer-implemented method of user authentication, the method comprising:

using a number of processors to perform the steps of:

combining, by a computer system, each of a number of user recurrent neural network with a system recurrent neural network to form respective unique combined recurrent neural networks, wherein each user recurrent neural network generates a unique respective user key and the system recurrent neural network generates a system key;

inputting, by the computer system, a same predetermined input into each combined recurrent neural network, wherein each combined neural circuit generates a unique combined key from the same input, and wherein each combined key differs from the system key and the user key of the user recurrent neural network comprising each combined recurrent neural network; and

associating, by the computer system, each combined key with a unique access authorization to authenticate a user.

12. A user authentication system, comprising:

a user recurrent neural network configured to generate a user key;

a computer system comprising:

a system recurrent neural network configured to generate a system key; and

a storage device configured to store program instructions and one or more processors operably connected to the storage device and configured to execute the program instructions to cause the computer system to:

combine the user recurrent neural network with the system recurrent neural network to form a unique combined recurrent neural network;

input a predetermined input into the combined recurrent neural network;

generate, with the combined recurrent neural network, a unique combined key from the input, wherein the combined key differs from both the user key and system key; and

associate the combined key with a unique access authorization to authenticate a user.

13. The user authentication system of claim 12 , further comprising inputting a second predetermined input to the combined recurrent neural network for multi-factor authorization.

14. The user authentication system of claim 13 , wherein the second predetermined input is supplied by the user.

15. The user authentication system of claim 14 , wherein the second predetermined input comprises one of:

a password;

a passphrase;

personal identification number;

voice recognition;

facial recognition;

retinal scan;

iris recognition;

fingerprints;

handprint; or

palm vein pattern.

16. The user authentication system of claim 13 , wherein the second predetermined input is provided by the computer system.

17. The user authentication system of claim 12 , wherein the user recurrent neural network and system recurrent neural network comprise liquid state machines.

18. The user authentication system of claim 12 , wherein the user recurrent neural network and system recurrent neural network comprise neuromorphic hardware.

19. The user authentication system of claim 12 , wherein the predetermined input comprises a pre-determined sequence of neuron spikes.

20. The user authentication system of claim 12 , wherein the combined recurrent neural network comprises a physically unclonable function.

21. The user authentication system of claim 12 , wherein the user recurrent neural network and system recurrent neural network combined produce encrypted output data as a function of input data and the structures of the recurrent neural networks.