Unhackable Random Number Generator Sidesteps Device Flaws for Ultimate Security

Researchers are continually striving to improve quantum random number generators (QRNGs), essential components of modern cryptography. Zhenguo Lu, Jundong Wu, and Yu Zhang, alongside Shaobo Ren, Xuyang Wang, and Hongyi Zhou et al. from Shanxi University and the Chinese Academy of Sciences, present a novel semi-device-independent QRNG designed to resist general attacks while considering practical limitations. This work is significant because it relaxes the stringent requirements for device characterisation typically demanded of QRNGs, only necessitating limits on the energy of emitted states. By leveraging the Kato inequality, the team demonstrates a protocol that generates more randomness than it consumes and achieves a net generation rate of 1.165 Mbps at 100MHz, offering a promising pathway towards robust security and high performance in practical QRNG systems.

Semi-device-independent quantum randomness via correlated variable inequalities

Scientists have developed a new quantum random number generator (QRNG) resistant to general attacks while maintaining a high generation rate. These advancements address a critical need for truly random numbers, essential for cryptography, simulations, and machine learning, where predictability is unacceptable.
Unlike conventional QRNGs reliant on meticulously characterised devices, this semi-device-independent (semi-DI) QRNG tolerates imperfections and even malicious manipulation of its components, offering a practical balance between security and performance. The research introduces a protocol that requires minimal characterisation of the source and measurement devices, limiting demands on real-world QRNG systems and broadening their potential applications.

This work leverages the Kato inequality for correlated variables to demonstrate that the protocol generates more randomness than it consumes, a key milestone in quantum information processing. Researchers implemented the scheme using a continuous-variable system with ternary inputs, employing heterodyne detection to compensate for phase fluctuations through data postprocessing.

This innovative approach alleviates the need for stringent system stability, simplifying experimental setup and enhancing practicality. The system achieves a net random number generation rate of 1.165 Mbps at 5.3×10^9 rounds, operating at 100MHz and demonstrating a significant improvement in both security and speed.

The core of this breakthrough lies in a multi-input semi-DI QRNG based on heterodyne detection, designed to withstand general attacks where adversaries can employ correlated measurement strategies. By treating the measurement apparatus as a ‘black box’, the protocol allows for potentially correlated quantum operations without imposing strict independence assumptions on the measurement devices.

The research constrains the energy of prepared quantum states, rather than requiring complete knowledge of them, further reducing the demands on practical implementation. This approach enables secure randomness extraction through data postprocessing of generation rounds, utilising a strong extractor to produce truly random numbers.

Furthermore, the use of heterodyne detection, which performs full phase-space tomography, eliminates the need for real-time phase stabilisation systems, streamlining the experimental setup and enhancing the protocol’s feasibility. The demonstrated capability of generating more randomness than consumed underscores the efficiency of this new QRNG design, offering a promising pathway towards robust security and high generation rates in a simplified experimental configuration. This work represents a significant step forward in the development of practical and secure quantum random number generators.

Ternary state preparation and statistical analysis for random number generation

A continuous-variable system employing heterodyne detection serves as the core of a new semi-device-independent quantum random number generator. The research implements a protocol utilising ternary inputs, where a user randomly designates each round as either a test round or a generation round with probabilities p t and 1-p t respectively.

During test rounds, the user prepares quantum states ρ x , where x belongs to the set {1, 2, 3}, and transmits them to a measurement station, recording the number of measurement outcomes y given input state ρ x as N y|x . The protocol’s security relies on constraining the overlap of quantum states across rounds by limiting the energy of the prepared states, ensuring |⟨ψ x |ψ x′ ⟩| ≥ 1, 2μ for x, x′ ∈ {1, 2, 3}, where μ represents the upper bound of the mean photon number.

This approach avoids rigorous characterisation of the source and measurement devices, demanding only a limit on the emitted states’ energy. The system operates at 100MHz, achieving a net random number generation rate of 1.165 Mbps after 5.3x 10 9 rounds, demonstrating a capacity to generate more randomness than it consumes.

Heterodyne detection enables phase compensation through data postprocessing, circumventing the need for a real-time phase stabilisation system and simplifying the experimental setup. The min-entropy is calculated based on the recorded counts N y|x , utilising established equations for both finite-size and asymptotic limits, allowing for secure random number extraction via postprocessing of data from the generation rounds. This work assumes an independent and identically distributed source, acknowledging potential classical correlations but preventing quantum correlations between the source, measurement devices, and the environment.

Ternary continuous-variable QRNG performance with heterodyne detection and min-entropy extraction

Researchers detail a semi-device-independent quantum random number generator (QRNG) achieving a net random number generation rate of 1.165 Mbps at 5.3×10^9 rounds. This system operates at 100MHz, demonstrating a practical balance between security and generation speed. The work focuses on a continuous-variable system employing ternary inputs, utilising heterodyne detection to compensate for phase shifts and reduce stringent system stability requirements.

The protocol incorporates a test-and-generation round approach, where test rounds randomly prepare states ρx, with x belonging to the set {1, 2, 3}, and generation rounds utilise a fixed state ρ3. Data from N rounds is recorded, tracking the count Ny|x representing measurement outcome y given input state ρx during test rounds.

The min-entropy is then calculated, enabling secure random number extraction via post-processing of data from the generation rounds using a strong extractor. The research constrains the lower bound of the overlap of quantum states by limiting the energy of the prepared states, ensuring |⟨ψx|ψx′⟩| ≥1 −2μ, where μ represents the upper bound of the mean photon number of the emitted states.

Assumptions include an independent and identically distributed source, and the independence of input random number generation from both devices and potential adversaries. Min-entropy evaluation under collective attacks in the asymptotic limit is performed, considering potential manipulation of the source and measurement apparatus by an adversary Eve.

The system’s security is further enhanced by addressing correlated quantum operations across multiple rounds, quantified through a stochastic process {Wn}. The expected value of this process, E(Wn|Fn−1), is calculated based on Eve’s quantum operation correlated with previous rounds, ensuring robustness against sophisticated attacks. This approach allows for quantification of private randomness loss, even with a finite number of rounds and potential inaccuracies in conditional probability estimation.

Practical high-speed quantum randomness from a heterodyne-detected source

A semi-device-independent quantum random number generator has been implemented utilising readily available commercial components. This protocol operates under relaxed assumptions, requiring only a verifiable energy bound on the source and permitting general adversarial attacks at the measurement stage.

Crucially, the security analysis incorporates finite-size effects, addressing a practical limitation of many quantum random number generators. The demonstrated system achieves a net random number generation rate of 1.165 Mbps when operating at 100MHz, representing a significant step towards practical, high-security QRNGs.

This is accomplished through the use of heterodyne detection, which enables phase compensation via data postprocessing, thereby reducing the need for complex system stabilisation. Compared to previous estimations of randomness, this work explicitly accounts for randomness overhead, demonstrating a positive net generation rate and enhancing feasibility.

Acknowledged limitations include the current repetition rate of the system, which impacts the overall generation speed. Future research will focus on increasing this repetition rate to further improve performance. The protocol also has potential for extension to encompass more input states and discretization outputs within continuous-variable systems, potentially increasing the minimum entropy per round and further enhancing the capabilities of the generator.