Entropy Data from Error Correction for Secure Random Number Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing security mechanisms in computing systems that rely on pseudo-random number generators are vulnerable to unauthorized reproduction of random numbers, which can compromise data protection, as the seed used to initialize these generators can be obtained by unauthorized entities.
Innovation Solution
The system enhances entropy in random number generation by incorporating error correction data from multiple memory modules, using a sideband processor to generate entropy data, and employing multiple layers of indirection, such as selective use of memory modules and scrambling, to make it difficult for attackers to derive the entropy data used for security operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If pseudo-random number generators are used for security operations, then random numbers can be generated for cryptographic operations and password generation, but the seed can be obtained by unauthorized entities leading to vulnerability
Solution Approach 1:
The patent introduces error correction data as an intermediary element between the original data and the random number generation process. By using error correction data from memory modules as entropy sources, the system creates an additional layer that unauthorized entities cannot easily access or reproduce, thereby securing the random number generation against seed compromise
Solution Approach 2:
The patent transitions from using traditional seed-based random number generation to a multi-dimensional entropy collection approach. Instead of relying on a single seed value, the system gathers entropy from multiple memory modules through error correction data, effectively moving from one-dimensional seed generation to multi-dimensional entropy aggregation, making reproduction infeasible
2Reliability
If error correction data from multiple memory modules is used to generate entropy, then the entropy of random number generation is significantly increased, but the device complexity increases
Solution Approach 1:
The patent makes the error correction circuit serve multiple functions: it not only corrects data errors but also generates entropy data for random number generation. By enabling the error correction circuit to perform both error correction and entropy generation, the system avoids adding separate dedicated hardware components, thus managing complexity while achieving high entropy
Solution Approach 2:
The error correction data is utilized by the system itself for entropy generation without requiring external entropy sources or additional complex hardware. The system leverages its existing error correction mechanisms to serve the dual purpose of data integrity and security, allowing the system to self-produce entropy using resources already present in the memory subsystem
Data Source
AI summary
In some examples, a system obtains error correction data for data stored in a memory, produces entropy data using the error correction data, adds the produced entropy data to a pool of entropy data, and performs a security operation using selected entropy data from the pool of entropy data.