Hybrid Random Number Generator for Cache Replacement Security
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage devices are vulnerable to power analysis attacks due to deterministic behavior in pseudo random number generators, which can be exploited to infer sensitive information from statistical analysis of physical signals like current consumption and electromagnetic radiation.
Innovation Solution
A secure data storage device is implemented using a combination of a physical random number generator based on a noise source and a pseudo random number generator, where the physical random number is used to re-initialize the pseudo random number generator and drive a replacer for selecting data storage units, allowing for a randomized replacement strategy that switches between deterministic and non-deterministic approaches based on random intervals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a pseudo random number generator is used for cache replacement strategy, then the implementation is simple and fast, but the deterministic behavior makes the system vulnerable to power analysis attacks
Solution Approach 1:
The patent combines a physical random number generator with a pseudo random number generator to create a hybrid system. The physical RNG introduces true randomness to break determinism and vulnerability to power analysis, while the pseudo RNG maintains speed and simplicity. This merging resolves the contradiction by integrating the security benefits of physical randomness with the performance advantages of pseudo-random generation.
Solution Approach 2:
The patent changes the fundamental parameter of randomness from pseudo-random to physically-generated random numbers. By using a physical random number generator that taps into environmental noise sources, the system transitions from deterministic behavior to truly non-deterministic behavior, thereby eliminating the vulnerability to power analysis attacks while maintaining acceptable performance.
2Object-affected harmful factors
If a physical random number generator is used to randomize replacement strategy, then security against power analysis attacks is improved, but the implementation complexity and cost increase
Solution Approach 1:
The patent merges a physical random number generator with a pseudo random number generator in a hybrid architecture. This combination allows the system to benefit from the security properties of physical randomness while using the pseudo-RNG to handle the bulk of replacement decisions, thereby reducing overall implementation complexity and cost compared to using only a physical RNG.
Solution Approach 2:
The patent applies physical random numbers only partially to the cache replacement strategy rather than exclusively. By using physical RNG for occasional or critical replacement decisions and pseudo-RNG for routine operations, the system achieves sufficient security improvement without the full complexity and cost burden of a complete physical RNG implementation.
3Object-affected harmful factors
If random replacement strategy is implemented, then vulnerability to power analysis attacks is reduced, but performance may be compromised compared to deterministic strategies
Solution Approach 1:
The patent changes the randomness parameter from pseudo-random to physical random numbers, providing true non-determinism that effectively counteracts power analysis attacks. This parameter change maintains cache performance by ensuring that replacement decisions are genuinely random rather than predictable, thereby preserving the performance advantages of random replacement strategies.
4Productivity
If deterministic replacement strategy is used, then performance is optimized and implementation is simple, but the system becomes vulnerable to statistical analysis of physical signals
Solution Approach 1:
The patent introduces dynamics into the replacement strategy by using a physical random number generator to randomly select between different replacement policies (e.g., LRU, random, FIFO). This dynamic switching prevents statistical analysis attacks by ensuring that the replacement behavior changes over time and cannot be predicted, while maintaining high cache performance through the use of effective replacement strategies.
Data Source
AI summary
A data storage device includes a plurality of data storage units, a physical random number generator with a noise source based on a physical noise process, for generating a random number, and a replacer for selecting a data storage unit wherein data is to be stored, depending on the random number. Selecting, on the basis of genuine random numbers, data storage units and/or lines to be replaced in the cache.