Quantum Codes Become More Efficient with Relaxed Design Constraints

Scientists Valentine Nyirahafashimana at the University Putra Malaysia, Kigali Independent University, Chouaıb Doukkali University and Canadian Quantum Research Center, have developed a new approach to quantum error correction that improves code performance by moving beyond strict geometrical constraints. They present a quasi-orthogonal framework for designing stabilizer codes, enabling controlled overlap between key code components while maintaining essential mathematical properties. This relaxation of traditional orthogonality expands the possibilities for code construction, yielding designs that achieve higher logical rates and improved performance at moderate distances. Their finite-length constructions demonstrate gains, up to two orders of magnitude improvement in logical error rates, fidelities, and trace distances under depolarizing noise with error rates up to 0.30, reflecting increased connectivity within the code’s structure and compatibility with existing decoding methods.

Stabilizer code performance surpasses orthogonal limits via quasi-orthogonal geometry

Logical error rates in stabilizer codes have improved by up to two orders of magnitude, a substantial leap beyond the capabilities of strictly orthogonal designs. Traditionally, quantum error-correcting codes (QEC) rely on orthogonal geometric constructions, where the supports of check operators, the components that detect errors, are strictly perpendicular in a mathematical space known as the binary symplectic space, denoted as $\mathbb{F}_{2}^{2}$. This orthogonality ensures that measuring one check operator does not inadvertently disturb the information encoded by others. However, this rigidity limits the design space and can hinder resource efficiency. With error rates up to 0.30 now achievable, this unlocks the potential for more robust quantum computations previously hampered by limitations in code performance. The new quasi-orthogonal geometric framework relaxes traditional constraints on code structure, permitting controlled overlap between code components while preserving essential quantum properties, specifically the symplectic commutation structure which is crucial for maintaining the code’s logical integrity.

This expansion of the design space enables codes approaching the Gilbert-Varshamov regime, a key benchmark for encoding efficiency. The Gilbert-Varshamov bound represents a theoretical limit on the achievable code rate, the ratio of logical qubits (the encoded information) to physical qubits (the hardware implementation). Approaching this regime signifies a highly efficient use of quantum resources. Finite-length constructions, including variants of the [[8,3,≈3]], [[10,4,≈3]] and others, consistently demonstrate these gains, reflecting increased connectivity and compatibility with existing decoding methods. The notation [[n,k,d]] represents a quantum code with n physical qubits, k logical qubits, and minimum distance d, which determines the code’s ability to correct errors. Analysis revealed reduced computational complexity and improved energy efficiency, with quantum error rates reaching 10−5 at 10 dB and 10−6 at 15 dB, alongside a bit-error-rate change of approximately 1.20 dB at 10−4. These figures indicate a significant improvement in signal-to-noise ratio and error resilience. These results suggest potential for implementation on near-term quantum hardware, where resource constraints are vital and efficient decoding is key. The reduced complexity is particularly important as decoding algorithms can be computationally demanding, often limiting the scalability of QEC schemes.

Controlled leakage enhances quantum error correction under depolarizing noise

Stabilizer codes underpin much of the current thinking around protecting fragile quantum information, but genuinely strong error correction demands pushing beyond established boundaries. A tantalising glimpse of what’s possible is offered by the new quasi-orthogonal framework, which deliberately introduces controlled ‘leakage’ between code components. This concept was previously avoided in favour of strict geometrical arrangements. The term ‘leakage’ refers to the small, controlled overlap between the supports of the X- and Z-check operators. While complete orthogonality prevents any interaction, this quasi-orthogonal approach allows for a limited degree of interaction, which, surprisingly, can enhance error correction performance under certain conditions. Currently, however, the benefits of this increased flexibility are demonstrated only under specific conditions, namely depolarizing noise up to an error rate of 0.30, leaving open the question of performance against other, potentially more realistic, error models. Depolarizing noise represents a common type of error in quantum systems, where quantum information is randomly lost.

Improved performance against even one error model represents a valuable step forward for quantum computers facing numerous potential error sources. Quantum systems are susceptible to a variety of noise sources, including bit-flip errors (where a qubit’s state is flipped from 0 to 1 or vice versa), phase-flip errors, and amplitude damping. This approach expands the set of tools for building more durable quantum systems, offering designers greater flexibility in optimising code structure and potentially achieving higher data throughput. Further investigation will focus on assessing the framework’s durability against more complex noise profiles, such as those incorporating both depolarizing and bit-flip errors. Understanding the interplay between different noise types and the quasi-orthogonal framework is crucial for developing truly robust QEC schemes.

Maintaining essential mathematical properties known as symplectic commutation while relaxing the requirement for strict orthogonality between code components creates a framework that expands the possibilities for code construction. Symplectic commutation ensures that the check operators commute with the encoded data, preserving the logical information. This enables the creation of codes with improved logical rates, bringing them closer to the Gilbert-Varshamov regime, a theoretical limit on encoding efficiency; this advancement is vital as quantum computers scale. The degree of controlled leakage can be tuned, offering a new parameter for optimising code performance and adapting to varying noise characteristics. This tunability allows for a more nuanced approach to error correction, potentially enabling the development of codes tailored to specific hardware platforms and noise environments. The ability to adjust the level of overlap between check operators provides a powerful tool for fine-tuning the code’s performance and resilience.

The research demonstrated improvements in quantum error correction using a new quasi-orthogonal geometric framework for stabiliser codes. By relaxing strict orthogonality constraints, while maintaining essential mathematical properties, the framework expands the possibilities for designing more efficient codes. Results showed logical error rates, fidelities, and trace distances improved by up to two orders of magnitude under depolarizing noise with error rates up to 0.30, compared to strictly orthogonal codes such as the [[8,3]], [[10,4]], [[13,1]], and [[29,1]] variants. Researchers intend to assess the framework’s performance against more complex noise profiles to further refine its robustness.