Decentralized RNG Architecture With Asynchronous Flow Engines
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Centralized random number generators (RNGs) in computing devices are vulnerable to attacks and can become a single point of failure, compromising the security of cryptographic keys and sensitive data.
Innovation Solution
Implement a decentralized multi-random number generator architecture using multiple independent RNG sources with asynchronous digital logic flow engines, including permutation, XOR, buffer, and zeroing operations, to generate and transform random numbers, which are then buffered for output to consumers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a centralized random number generator is used, then the device structure is simple, but the security is compromised due to single point of failure vulnerability
Solution Approach 1:
The patent divides the centralized RNG into multiple independent RNG sources (first RNG, second RNG, etc.), each generating random numbers that are then combined through flow engines. This segmentation eliminates the single point of failure vulnerability while maintaining system security.
Solution Approach 2:
The patent merges outputs from multiple independent RNG sources through cryptographic flow engines to produce a unified random number output. This combining approach maintains the security benefits of decentralization while providing a single interface to consumers.
2Reliability
If multiple independent RNG sources are used, then security is enhanced, but the device complexity increases
Solution Approach 1:
The patent introduces flow engines as intermediary components that process and combine random numbers from multiple RNG sources. These flow engines implement cryptographic transformations (permutation, XOR, buffer, zeroing operations) to securely merge the decentralized sources while managing the complexity of coordinating multiple RNGs.
3Reliability
If flow engines are used to transform random numbers, then entropy is maintained, but power consumption increases
Solution Approach 1:
The patent implements flow engines that operate periodically rather than continuously, processing random numbers through permutation, XOR, buffer, and zeroing operations only when needed. This periodic operation maintains entropy requirements while reducing overall power consumption compared to continuous processing.
Data Source
AI summary
Systems and techniques are provided for random number generation. For instance, a process can include generating, using a first random number generator of a plurality of random number generators, a first random number; inputting the first random number to a first flow engine of a plurality of flow engines, wherein the first flow engine is coupled to a random number consumer, of one or more random number consumers, through a first set of flow engines, of the plurality of flow engines, wherein the plurality of flow engines are unclocked; randomly transforming, by the first set of flow engines, the first random number to a second random number; and buffering the second random number from the first set of flow engines for output to the random number consumer.