Secure authentication using recurrent neural networks
| DWPI Title: Method for authenticating users using unique combined keys generated by chaotic neural networks in client devices, involves generating unique combined key from input, and associating combined key with unique access authorization to authenticate user |
| Abstract: A computer-implemented method of user authentication is provided. The method comprises combining, by a computer system, a user recurrent neural network with a system recurrent neural network to form a unique combined recurrent neural network. 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. The computer system inputs a predetermined input into the combined recurrent neural network, and the combined recurrent neural network generates a unique combined key from the input, wherein the combined key differs from both the user key and system key. The computer system then associates the combined key with a unique access authorization to authenticate a user. |
| Use: Method for authenticating users using unique combined keys generated by chaotic neural networks in client devices. Uses include but are not limited to computers, workstations, network computers, mobile phones, tablet computers and smart glasses. |
| Advantage: The method enables ensuring that a combination key yields a neuron activity that would be unlikely or impossible to observe resulting in a neural computational physically unclonable function (PUF) that offers increased security and more flexible utility than previous electronic PUFs. |
| Novelty: The method (800) involves combining (802) a user recurrent neural network with a system recurrent neural network to form a unique combined recurrent neural network by a computer system, where the user recurrent neural network generates a unique user key, and the system recurrent neural network generates a system key. A predetermined input is input (804) into the combined recurrent neural network. A unique combined key from the input is generated (806), where the combined key differs from both the user key and system key. The combined key with a unique access authorization is associated (808) to authenticate a user, where the predetermined input comprises a pre-determined sequence of neuron spikes. |
| Filed: 3/9/2021 |
| Application Number: US17196295A |
| Tech ID: SD 15319.0 |
| This invention was made with Government support under Contract No. DE-NA0003525 awarded by the United States Department of Energy/National Nuclear Security Administration. The Government has certain rights in the invention. |
| Data from Derwent World Patents Index, provided by Clarivate All rights reserved. Republication or redistribution of Clarivate content, including by framing or similar means, is prohibited without the prior written consent of Clarivate. Clarivate and its logo, as well as all other trademarks used herein are trademarks of their respective owners and used under license. |