High speed self-testing quantum random number generation without detection loophole

Yang Liu, Xiao Yuan, Minghan Li, Weijun Zhang, Qi Zhao, Jiaqiang Zhong, Yuan Cao, Yu-Huai Li, Luo-Kan Chen, Hao Li, Tianyi Peng, Yu-Ao Chen, Cheng-Zhi Peng, Sheng‐Cai Shi, Zhen Wang, Lixing You, Xiongfeng Ma, Jingyun Fan, Qiang Zhang, Jian-Wei Pan2017 · DOI

TL;DR

Researchers have developed a new method for generating truly random numbers using quantum mechanics, achieving a record speed of 114 bits per second. This "self-testing" approach guarantees the randomness without needing to trust the devices used, a crucial step for secure communication and fundamental physics experiments.

What Problem Does This Paper Address?

Generating truly random numbers is essential for many applications, from secure cryptography to scientific simulations. While classical computers rely on deterministic algorithms that can, in principle, be predicted, quantum mechanics offers a source of genuine randomness. However, practical quantum random number generators (QRNGs) face challenges:

  • Device Trust: Adversaries might gain information about or even manipulate the devices generating the randomness, compromising its security.
  • Loopholes: Experimental setups can have "loopholes" (like detection or locality loopholes) that allow classical explanations for observed correlations, undermining the quantum nature of the randomness.
  • Speed: Previous methods for generating provably random numbers were often too slow for practical applications.

This paper tackles these issues by demonstrating a "self-testing" QRNG that is independent of implementation details and closes the detection loophole, all while achieving a significantly higher generation rate.

Methods

The core of this work is a loophole-free Bell test using entangled photons. Here's how it works:

  • Entangled Photons: Pairs of photons are created in a special quantum state where their properties are linked, no matter how far apart they are.
  • Alice and Bob: Two parties, Alice and Bob, each receive one photon from an entangled pair. They independently choose measurement settings (like the angle to measure polarization) and record their outcomes.
  • Bell Test: The experiment measures the correlations between Alice's and Bob's outcomes. If the correlations are stronger than what classical physics allows (as dictated by Bell's inequality), it proves that the outcomes are genuinely random and cannot be explained by local hidden variables.
  • Self-Testing: The key innovation is that the violation of Bell's inequality itself acts as a "test" for the randomness. The stronger the violation, the more random the output bits are guaranteed to be, regardless of how the devices were built or if they were tampered with. This is achieved by analyzing the CHSH game value.
  • Randomness Extraction: After the Bell test confirms the quantum nature of the randomness, a process called Toeplitz-matrix hashing is used to extract a final string of random bits. This process amplifies the randomness and ensures its uniformity.

Datasets

Not discussed in the paper. The experiment generates its own data through the Bell test.

Key Insights and Ideas

  • Device Independence is Key: The "self-testing" aspect is a major breakthrough. It means we don't need to trust the complex internal workings of the quantum devices; the randomness is certified by the observed correlations themselves.
  • Closing the Detection Loophole: By ensuring high photon detection efficiency, the experiment rules out explanations where the randomness is faked by manipulating detection probabilities.
  • Speed-Up: Achieving 114 bits/s is a significant improvement over previous self-testing QRNGs, making it more viable for real-world applications.
  • Theoretical Foundation: The work builds on advanced theoretical frameworks like the entropy accumulation theorem, which allows for robust security analysis even without assuming independent and identically distributed (i.i.d.) outcomes.

Key Results

  • Random Bit Rate: The experiment achieved a final random bit rate of 114 bits/s.
  • Failure Probability: The generated random bits have a failure probability (meaning they might not be perfectly random) of less than 10⁻⁵.
  • Bell Inequality Violation: The experiment demonstrated a significant violation of Bell's inequality, confirming the quantum nature of the generated randomness.
  • High Detection Efficiency: The photon detection efficiencies were 78.6% for Alice and 80.2% for Bob, which is crucial for closing the detection loophole.

Limitations and Future Work

  • Speed vs. Optimal: While 114 bits/s is a record for self-testing QRNGs, it's still significantly lower than the theoretical maximum achievable with i.i.d. assumptions. The paper notes that the obtained randomness is about 60% of the optimal asymptotic value.
  • Further Speed Improvements: The authors suggest that further increasing the operation repetition rate and upgrading technology could lead to even faster randomness generation.
  • Cosmic RNGs: A prospective direction is to use cosmic RNGs in loophole-free CHSH games to generate "genuinely quantum-certified" random bits.
  • Maximum Extractable Randomness: An open question remains about the absolute maximum extractable randomness from a CHSH game or general nonlocal games.

Summary updated

Abstract

We experimentally demonstrate a self-testing quantum random number generation based on an detection-loophole free Bell test with entangled photons. We achieve a final random bit rate of 114 bits/s, with a failure probability less than 10−5.
BibTeX
@misc{liu2017high,
  author = {Yang Liu and Xiao Yuan and Minghan Li and Weijun Zhang and Qi Zhao and Jiaqiang Zhong and Yuan Cao and Yu-Huai Li and Luo-Kan Chen and Hao Li and Tianyi Peng and Yu-Ao Chen and Cheng-Zhi Peng and Sheng‐Cai Shi and Zhen Wang and Lixing You and Xiongfeng Ma and Jingyun Fan and Qiang Zhang and Jian-Wei Pan},
  title = {{High speed self-testing quantum random number generation without detection loophole}},
  year = {2017},
  doi = {10.1364/fio.2017.fth2e.1},
}

Loading related content...

Figure