Modular Reduction Hardware Using Randomized Quotient Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cryptographic algorithms using large-integer multiplication and modular reduction are vulnerable to power analysis and timing attacks, and current methods for efficient modular reduction, such as the Barrett method, do not adequately balance security and speed.
Innovation Solution
A computer-implemented modular reduction method that systematically underestimates the quotient with a randomized error, using a random number generator to inject error into the quotient estimation, ensuring faster computations and enhanced security against cryptoanalysis attacks while maintaining mathematical equivalence with true results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional modular reduction methods (Barrett, Quisquater, Montgomery) are used, then computation speed is improved, but security against power analysis and timing attacks deteriorates
Solution Approach 1:
The invention introduces dynamic randomization into the quotient estimation process by adding a randomized error term to the estimated quotient. This makes the computational path variable and unpredictable, preventing attackers from analyzing power consumption patterns or timing to deduce cryptographic keys, while still maintaining computational efficiency through the use of approximate quotient estimation.
Solution Approach 2:
The invention changes the parameter of quotient estimation from an exact value to an approximate value with controlled error. By systematically underestimating the quotient with a randomized error of a few bits, the method achieves faster computation while the randomized variation prevents cryptographic analysis attacks.
2Measurement precision
If exact quotient computation is performed, then accuracy is improved, but computation speed deteriorates
Solution Approach 1:
The invention applies partial action by computing only an approximate quotient rather than the exact quotient. The estimated quotient with randomized error is sufficient for the modular reduction to produce a congruent remainder, eliminating the need for expensive exact division while maintaining cryptographic correctness.
Solution Approach 2:
The method changes the accuracy parameter of quotient computation from exact to approximate with controlled error bounds. This parameter change enables faster computation using multiplication and shifting operations instead of full division, while the randomized error ensures security against analysis attacks.
3Reliability
If randomized error is injected into quotient estimation, then security is improved, but computation accuracy deteriorates
Solution Approach 1:
The invention converts the potential harm of inaccuracy into a benefit by using randomized error. The inexact quotient estimation would normally produce incorrect remainders, but by carefully controlling the error bounds and using the property of modular congruence, the method produces remainders that are congruent to the true remainder, making the inaccuracy beneficial for security while maintaining mathematical correctness.
Data Source
AI summary
A cryptographically secure, computer hardware-implemented modular reduction method systematically underestimates and randomizes an approximate quotient used for computation of a remainder. The randomizing error injected into the approximate quotient is limited to a few bits, e.g. less than half a word. The computed remainder is congruent with but a small random multiple of the residue, which can be found by a final set of subtractions by the modulus. In addition to a computational unit and operations sequencer, the computing hardware also includes a random or pseudo-random number generator for producing the random error. The modular reduction method thus resists hardware cryptoanalysis attacks, such as timing and power analysis attacks.