[Paper Review] Timing Attack Resilient Decoding Algorithms for Physical Unclonable Functions

Sven Puchinger, Sven Müelich+2 more|arXiv (Cornell University)|Nov 23, 2016

Chaos-based Image/Signal Encryption3 citations

TL;DR

This paper proposes timing attack-resilient decoding algorithms for Physical Unclonable Functions (PUFs) using list decoding of Reed–Solomon codes in a concatenated coding scheme. By extending the decoding radius beyond half the minimum distance, the approach achieves lower block error rates than prior schemes while ensuring constant-time execution through permutation-based masking and algorithmic hardening, enabling secure key reproduction under side-channel threats.

ABSTRACT

This paper deals with the application of list decoding of Reed--Solomon codes to a concatenated code for key reproduction using Physical Unclonable Functions. The resulting codes achieve a higher error-correction performance at the same code rate than known schemes in this scenario. We also show that their decoding algorithms can be protected from side-channel attacks on the runtime both by masking techniques and by directly modifying the algorithms to have constant runtime.

Motivation & Objective

• To improve error-correction performance in PUF-based key reproduction systems beyond existing schemes.
• To protect decoding algorithms from timing side-channel attacks that exploit variable runtime.
• To achieve information-theoretic security against side-channel leakage via permutation-based masking of Hamming distance computations.
• To enable practical deployment of high-performance PUFs in real-world hardware with side-channel resistance.
• To lay the foundation for integrating advanced soft-decision decoding like Kötter–Vardy into PUF systems.

Proposed method

• Uses list decoding of Reed–Solomon codes in a concatenated coding scheme to extend the decoding radius beyond half the minimum distance, improving error-correction capability.
• Employs a random permutation of codewords to mask the order of Hamming distance computations, ensuring constant-time execution regardless of the actual codeword.
• Applies the masking technique from [7] to the list decoding process, proving it provides information-theoretic security against timing attacks.
• Defines a permuted list of codewords and computes Hamming distances to the received word in permuted order, preserving uncertainty about the true codeword.
• Uses the property that adding a codeword to all codewords and the received word preserves Hamming distances under permutation, ensuring indistinguishability.
• Introduces a formal proof that the conditional entropy of the codeword remains unchanged even if an attacker observes the ordered list of Hamming distances.

Experimental results

Research questions

• RQ1 Can list decoding of Reed–Solomon codes improve the block error rate performance in PUF-based key reproduction compared to conventional decoding methods?
• RQ2 Can the decoding algorithm be made resistant to timing side-channel attacks without sacrificing performance?
• RQ3 Does masking the order of Hamming distance computations preserve information-theoretic security in the presence of timing leakage?
• RQ4 Can the proposed constant-time decoding method be integrated into existing PUF architectures with minimal overhead?
• RQ5 What is the theoretical guarantee of security when an attacker observes the ordered list of Hamming distances to all codewords?

Key findings

• The proposed list decoding scheme achieves a lower block error probability than existing codes and decoders in the PUF scenario, due to the extended decoding radius beyond half the minimum distance.
• The masking technique based on random permutation of codewords ensures that the conditional entropy of the codeword remains unchanged even if the attacker observes the ordered list of Hamming distances.
• The proof demonstrates that the uncertainty about the true codeword does not decrease under the observation of permuted Hamming distances, confirming information-theoretic security.
• The decoding algorithm can be made constant-time by processing codewords in a permuted order, eliminating timing side-channel leakage.
• The approach is compatible with advanced decoding algorithms such as the Kötter–Vardy soft-decision decoder, which may further improve performance.
• The method provides a foundation for integrating high-performance, side-channel-secure decoding into practical PUF-based key management systems.