New protocol cuts quantum hardware needs for secure computing

Researchers at Quandela, a company based in Massy, France, have developed a new protocol that drastically reduces the technological demands on both clients and servers within Secure Delegated Quantum Computation (SDQC). Existing SDQC protocols often require servers to perform complex operations like isolating single photons and performing polarisation-preserving photon-number quantum non-demolition measurements.

This new approach bypasses those hurdles by relying on interacting quantum emitters capable of generating spin-photon entanglement. The protocol allows a client to manipulate only an attenuated laser pulse while maintaining information-theoretic composable security, hinting at readiness for experimental implementation given recent demonstrations of large entangled photonic states.

Quandela’s Protocol Reduces Quantum Hardware for Secure Computing

This represents a significant reduction in the quantum technological capabilities traditionally required of the client in Secure Delegated Quantum Computation (SDQC) schemes, which often demand single-qubit sources or measurements. Existing SDQC protocols aim to enable users to use remote quantum servers without compromising data security or algorithm integrity, but frequently impose substantial hardware demands on the user. The Quandela protocol instead relies on the server handling interacting quantum emitters structured to generate spin-photon entanglement, a potentially more feasible configuration for current and near-future quantum hardware.

This simplification addresses a key limitation of earlier SDQC designs, which often struggled with the practical challenges of implementing these detailed server-side procedures. The security of this protocol is not merely an added benefit. It is “information-theoretic and composable in the Abstract Cryptography framework,” according to the research.

This level of security ensures that even if portions of the protocol are reused in other cryptographic constructions, the overall security guarantees remain intact, a feature lacking in some previous designs. Composability is a critical aspect of modern cryptographic design, as it allows for the building of complex, secure systems from simpler, well-understood components. The protocol uses Measurement-Based Quantum Computing (MBQC), a method that begins with a large entangled state, known as a graph state, and performs computation by adaptively measuring parts of it.

This approach allows for blind computation, where the server executes the client’s algorithm without learning anything about the input, the algorithm itself, or the final result. The team built upon an existing efficient protocol for Blind Delegated Quantum Computing (BDQC) with weak coherent pulses, extending it to create a fully secure SDQC protocol.

The researchers demonstrate that their protocol can delegate classical input and output BQP computations with an exponentially-low security error, while still maintaining the minimal hardware requirements for both client and server. This combination of security and practicality is a key advancement, as previous protocols often forced a trade-off between the two. The client’s role is limited to sending weak coherent pulses, while the server manages the entangled quantum emitters, streamlining the overall process.

The team analyzed the experimental feasibility of their protocol, specifically examining the trade-off between correctness and security in the presence of photon loss after emission. This analysis is important for assessing the protocol’s viability in real-world scenarios, where imperfections in hardware and transmission can significantly impact performance. The study considered the effects of photon loss, a common challenge in photonic quantum computing, and developed strategies to mitigate its impact on both security and accuracy.

The work builds on existing research in delegated quantum computing, but addresses shortcomings in previous approaches. The authors note that other protocols do not resolve issues stemming from complex server operations, making their protocol more suitable for implementation. The development of this protocol represents a step toward a more practical and scalable architecture for quantum information processing.

By reducing the technological demands on both the client and the server, it opens up the possibility of wider adoption of SDQC, enabling more users to benefit from the power of remote quantum computation without requiring significant investment in quantum hardware. This could accelerate the development of quantum applications in fields such as drug discovery, materials science and financial modeling.

Secure Delegated Quantum Computation: Protecting Client Data & Algorithms

This new protocol achieves information-theoretic and composable security while allowing the client to operate using only an attenuated laser pulse, simplifying the client’s technological footprint considerably. This approach allows for encrypted graph states to be created, where the client prepares encrypted qubits, now realized as attenuated laser pulses, and sends them to the server. Security within the protocol is bolstered by a phase-randomisation technique applied to the attenuated laser pulses, with the polarisation angle of each pulse serving as a secret key for encrypting the computation.

The probability of both incorrect state generation and information leakage to the server decreases exponentially with the number of pulses sent, a relationship proven within a composable security framework. This exponential decrease is a key element in establishing the protocol’s robustness against potential attacks and ensures the confidentiality of the client’s data and algorithm.

The team’s analysis extends beyond theoretical security, examining the practical feasibility of the protocol in the face of real-world imperfections. Specifically, they investigated the trade-off between maintaining computational correctness and ensuring security when photon loss occurs after emission. The researchers demonstrated that the protocol can function effectively even with a degree of photon loss, provided sufficient measures are taken to mitigate its effects. The culmination of this work is the formalization of a theorem establishing the existence of an efficient, composably secure BDQC protocol.

“There exists an efficient composably secure BDQC protocol for classical inputs with an exponentially-low security error in which the client only sends weak coherent pulses and the server only manipulates interacting quantum emitters,” the authors state. The SDQC protocol’s ability to detect deviations from the client’s instructions is a critical feature, as it allows the client to verify the computation’s outcome with high probability.

This verification step is essential for building trust in the delegated computation and ensuring that the results are accurate and reliable. The protocol’s design prioritises minimising the technological burden on both parties, making it a potentially scalable solution for secure quantum computing. The team’s work demonstrates a pathway toward a future where users can use remote quantum servers without compromising the security of their data or algorithms.

The study’s findings suggest a significant advancement in the field of secure quantum computation, offering a practical approach that addresses limitations in existing protocols. The protocol’s reliance on interacting quantum emitters, rather than complex single-photon manipulation, positions it as a promising candidate for implementation on near-term quantum hardware.

Semi-Classical Light Enables Client Use of Attenuated Laser Pulses

Existing approaches to secure delegated quantum computation often demand substantial technological resources from either the client or the server, but a new protocol minimizes these demands by using attenuated laser pulses and interacting quantum emitters. Unlike many predecessors, this protocol does not require the client to manipulate individual qubits, instead relying on a phase-randomized attenuated laser source and standard waveplates for control. This simplification significantly lowers the barrier to entry for end-users seeking to use remote quantum processing power without possessing advanced quantum hardware.

The server’s requirements are also streamlined; the protocol circumvents the need for photon-number quantum non-demolition measurements and deterministic photon-photon gates previously considered essential. This design choice is a key element in achieving information-theoretic composable security, a standard that guarantees protection against even a malicious server or eavesdropper.

The protocol’s security has been formally proven within the Abstract Cryptography framework, ensuring its robustness against sophisticated attacks. A central innovation lies in the emulation of single-qubit transmissions using only the attenuated laser pulses, a technique that addresses a key challenge in previous semi-classical approaches. The client encodes a secret key within the polarization of each pulse. The server applies corrections based on the client’s chosen angles and measurement results, as detailed in Protocol 1.

While attenuated laser pulses inherently leak more information about the secret key than single qubits, the protocol accounts for this through a redundant encoding of the single qubit state, effectively mitigating the risk. Recent experimental advances in generating large entangled photonic states, including the creation of the largest such state to date, suggest the feasibility of implementing this protocol with current technology.

The protocol’s efficiency is further enhanced by its ability to create encrypted graph states, a fundamental resource for universal blind quantum computation. The client instructs the server to measure qubits within this encrypted graph state, simultaneously undoing the encryption and performing the desired computation. The team’s work demonstrates a clear advantage over earlier methods, such as the protocol from Dunjko et al., which necessitated both photon-number quantum non-demolition measurements and photon-photon entangling gates.

A recent protocol proposed by Takeuchi et al. also explored semi-classical light for secure delegated quantum computation, but it shares similarities with earlier approaches in its reliance on complex server-side operations. The new protocol, however, diverges significantly by avoiding these hurdles, offering a more streamlined and potentially more practical solution.

The privacy amplification technique employed is similar to those used by Dunjko et al., but the core advantage lies in the elimination of the need for non-destructive photon measurements and deterministic entangling gates. The researchers detail in Section 2 how to generate graph states from quantum emitters and weak coherent pulses, laying the groundwork for the secure computation process. This pragmatic approach acknowledges the challenges of real-world implementation and provides insights into potential improvements.

“Our scheme is thus not only secure but also experiment-ready,” the researchers state, highlighting the protocol’s potential for near-term deployment. The team’s work represents a step toward a future where secure quantum computation is accessible to a wider range of users, reducing the technological burden on both clients and servers.

Limitations of Existing Secure Delegation Protocols

Existing secure delegated quantum computation (SDQC) protocols often place significant demands on the client, requiring either a source of qubits or the ability to perform single-qubit measurements, limiting their accessibility for users with restricted quantum resources. This reliance on client-side quantum capabilities contrasts sharply with the potential for widespread adoption, as it presupposes a level of technological sophistication not yet universally available. The need for specialized hardware on the client side introduces a barrier to entry for many potential users of delegated quantum computing services.

These server-side protocols necessitate the execution of large, cryptographically-secure functions to generate encrypted qubits, demanding considerable processing power and memory. The security assurances offered by these purely classical protocols are not information-theoretically provable, instead relying on the presumed difficulty of certain computational problems, a foundation that may not withstand advances in quantum algorithms. This reliance on computational hardness also prevents their use as building blocks in larger, composable cryptographic constructions.

The limitations extend to the server’s hardware requirements as well. Many SDQC protocols demand operations that are difficult to realize with current technology. These complex operations increase the cost and complexity of building and maintaining a quantum server capable of supporting secure delegation, hindering practical implementation. The difficulty of these tasks has historically been a significant impediment to the widespread deployment of SDQC systems.

The development of protocols that address these limitations is important for realizing the full potential of SDQC. Information-theoretic security relies solely on the laws of physics, providing a stronger foundation for trust, while composability ensures that the security of a protocol remains intact even when used as a component in a larger cryptographic system. Protocols lacking these features may be vulnerable to unforeseen attacks or may not be suitable for complex applications requiring layered security.

The ability to generate graph states is a fundamental requirement for many SDQC protocols, and the methods used to achieve this can significantly impact the overall feasibility of the system. Existing approaches often rely on complex quantum emitters and precise control of weak coherent pulses.

The efficiency and reliability of these methods are critical for achieving high success rates in the delegated computation. A robust and scalable method for generating graph states is therefore essential for practical SDQC. Analysis of protocol performance must also account for realistic hardware limitations, particularly photon loss after emission, and photon loss introduces errors into the computation and reduces the overall success probability. Understanding the impact of these errors and developing techniques to mitigate them is important for building a reliable SDQC system.

A pragmatic approach to error analysis, considering the specific characteristics of available hardware, is essential for guiding the development of practical protocols. A successful protocol will not only be secure but also practical, meaning that it can be implemented with existing or near-future technology.

More like this

Quantum SecurityQuantum XChange warns of AI security gap amid rising attacks

Quantum SecurityIDEMIA gets first EMVCo certification for quantum-safe payment cards

Quantum HardwareQUTE Electronics signs deal for Andhra Pradesh’s quantum hub

Quantum HardwareHelios says high-fidelity qubits beat fast gates for real problems