GAN Encryption Key Generation for High-Entropy Cryptography

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Traditional methods for generating cryptographic keys suffer from vulnerabilities due to low entropy and predictability, making them susceptible to attacks, and consume excessive computing resources.

Innovation Solution

A GAN model is employed to generate encryption keys with maximum entropy by iteratively refining a key generator and discriminator network, enhancing randomness and unpredictability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional pseudo-random processes are used to generate encryption keys, then the generation process is simple and fast, but the entropy and unpredictability of the keys are low, making them vulnerable to attacks

Engineering Contradiction:

Improvesecurity of encryption keysVSAvoidcomplexity of key generation system

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

A GAN model is introduced as an intermediary between the key generation process and the security requirement. The generator network creates candidate keys while the discriminator network evaluates their entropy and randomness, mediating the transformation from simple pseudo-random processes to high-entropy cryptographic keys without requiring complex manual intervention

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

Traditional mechanical or algorithmic pseudo-random number generators are replaced with a neural network-based GAN system. The generator network learns to produce high-entropy keys through training rather than relying on predetermined algorithms, substituting a data-driven approach for traditional cryptographic primitive

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If traditional key generation methods are used, then the system is simple to implement, but excessive computing resources are consumed due to vulnerability to attacks and need for repeated generation

Engineering Contradiction:

Improvecomputing resource efficiencyVSAvoidsecurity against attacks

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The GAN model is trained in advance on a dataset of high-entropy keys to learn the characteristics of secure key generation. This preliminary training phase enables the generator to produce secure keys directly during operation without requiring repeated attempts or validation, improving computing resource efficiency during actual key generation

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The GAN system is self-validating through the discriminator network, which automatically evaluates the entropy and quality of generated keys without requiring external verification. This self-service mechanism eliminates the need for repeated generation and external security audits, reducing computing resource consumption

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS12556383B2Systems and methods for utilizing machine learning models to generate encryption keys

Publication Date: 2026.02.17 VERIZON PATENT & LICENSING INC

AI summary

A device may receive input key material, and may process the input key material, with a trained generative adversarial network (GAN) model, to generate an encryption key with a maximized entropy. The trained GAN model may include a key generator network model trained to generate encryption keys that generalize key derivation functions with higher entropy to enhance cryptographic security, and a key discriminator network model trained to predict authenticities of the encryption keys generated by the key generator network model. The device may perform one or more actions based on the encryption key.