[Paper Review] A modular framework for randomness extraction based on Trevisan's construction

Wolfgang Mauerer, Christopher Portmann+1 more|arXiv (Cornell University)|Dec 3, 2012

Complexity and Algorithms in Graphs25 references19 citations

TL;DR

This paper presents a high-performance, modular implementation of Trevisan's randomness extractor, enabling efficient and secure extraction of uniform randomness from weakly random sources using a short seed. It improves upon prior work by achieving orders-of-magnitude higher throughput, scaling to realistic data sizes, and providing explicit non-asymptotic bounds for quantum-secure privacy amplification in quantum key distribution systems.

ABSTRACT

Informally, an extractor delivers perfect randomness from a source that may be far away from the uniform distribution, yet contains some randomness. This task is a crucial ingredient of any attempt to produce perfectly random numbers---required, for instance, by cryptographic protocols, numerical simulations, or randomised computations. Trevisan's extractor raised considerable theoretical interest not only because of its data parsimony compared to other constructions, but particularly because it is secure against quantum adversaries, making it applicable to quantum key distribution. We discuss a modular, extensible and high-performance implementation of the construction based on various building blocks that can be flexibly combined to satisfy the requirements of a wide range of scenarios. Besides quantitatively analysing the properties of many combinations in practical settings, we improve previous theoretical proofs, and give explicit results for non-asymptotic cases. The self-contained description does not assume familiarity with extractors.

Motivation & Objective

• To provide a practical, high-performance implementation of Trevisan's randomness extractor that supports real-world applications in quantum cryptography and simulations.
• To enable scalability to large data sets—exceeding prior work by 10 orders of magnitude—where the extracted randomness surpasses the seed size.
• To offer a modular framework allowing flexible combination of building blocks to tailor the extractor to specific application requirements.
• To improve and clarify theoretical proofs of quantum security, providing explicit non-asymptotic bounds for the extractor's performance.
• To serve as a self-contained guide for experimentalists needing to deploy extractors without deep theoretical background.

Proposed method

• The framework implements Trevisan’s extractor using a modular architecture, where different components—such as one-bit extractors and weak designs—can be flexibly combined.
• It employs a weak $(m,t,r,d)$-design to structure the seed usage, ensuring low overlap between subsets and enabling strong extractor properties.
• The construction uses a one-bit quantum-proof extractor as a base, which is then composed via Trevisan’s method to yield a full extractor with improved security and efficiency.
• The weak design is constructed using finite fields and recursive partitioning, with parameters tuned to minimize overlap and ensure $r=1$ or $r<2e$.
• Theoretical analysis proves that the resulting extractor is a quantum-proof $(k+rm+\log 1/\varepsilon, 3m\sqrt{\varepsilon})$-strong extractor, with explicit non-asymptotic bounds.
• The implementation is optimized for performance, achieving throughput several orders of magnitude higher than prior implementations such as Ma et al. (2012).

Experimental results

Research questions

• RQ1 How can Trevisan’s randomness extractor be efficiently implemented in practice for large-scale data sets?
• RQ2 What modular design choices enable high performance and flexibility in extractor deployment across diverse applications?
• RQ3 How do the non-asymptotic parameters of the extractor affect its security and efficiency in quantum-secure settings?
• RQ4 Can the theoretical quantum security of Trevisan’s extractor be concretely quantified and verified in practical implementations?
• RQ5 How does the performance of this implementation compare to prior work in terms of throughput and scalability?

Key findings

• The implementation achieves throughput several orders of magnitude higher than the prior work by Ma et al. (2012), enabling real-time processing of large data sets.
• The framework successfully scales to data sets exceeding 10 orders of magnitude larger than those considered in Ma et al. (2012), where extracted randomness surpasses the seed size.
• Theoretical analysis confirms that the extractor is a quantum-proof $(k+rm+\log 1/\varepsilon, 3m\sqrt{\varepsilon})$-strong extractor, with explicit non-asymptotic bounds.
• The weak design construction ensures $r=1$ or $r<2e$, minimizing overlap and enabling strong extractor behavior with minimal seed correlation.
• The full source code is released under the GPL v2, enabling free use, modification, and integration into research and commercial systems.
• The implementation demonstrates that Trevisan’s extractor outperforms two-universal hashing in regimes where the extracted randomness exceeds the seed size, validating its theoretical advantage.