Researchers Design Bosonic Codes with Finite Stellar Rank for Constrained Quantum Error Correction

Rui Wang at Chalmers University of Technology and colleagues have achieved a breakthrough in bosonic quantum error correction, a method of protecting quantum information using particles of light, like building a shield around a fragile object. Stellar rank is a measure of how complex a quantum state is to create, similar to the number of ingredients and steps needed for a complicated recipe, and this low rank signifies fewer resources are needed than previously believed.

Bosonic code performance improves with stellar rank optimisation for enhanced qubit protection

This breakthrough signifies that logical qubits can be protected using fewer resources than previously thought, representing a substantial improvement in the field. By optimising bosonic encodings directly at a fixed stellar rank, grid-like structures emerge under photon loss, and approximately rotation-symmetric encodings are favoured under dephasing, establishing stellar rank as a valuable metric for evaluating and improving bosonic quantum error correction. A stellar rank of two now exceeds the break-even point for all levels of dephasing, a type of quantum noise that scrambles information. Stellar fidelity quantified the quality of these approximations, revealing that high fidelity alone does not automatically translate to better error correction performance.

Optimising bosonic codes via stellar rank reveals resource demands for quantum error correction

Researchers at Chalmers University of Technology, PSL University, and The University of Sydney has demonstrated a pathway towards more practical bosonic quantum error correction, but their current work relies on specific code families, cat and Gottesman-Kitaev-Preskill codes. This leaves open whether these findings generalise to other, potentially more efficient, encoding schemes. While the team successfully optimised codes at a fixed stellar rank, a measure of state complexity, the study acknowledges that scaling these optimised designs to larger, more complex quantum systems presents a significant challenge.

Nevertheless, even with limitations regarding generalisation to all quantum encoding methods, this research provides valuable insight into designing practical bosonic quantum error correction systems. The team’s focus on stellar rank establishes a concrete link between resource demands and code performance under realistic conditions, specifically photon loss and dephasing. Optimisation of quantum codes at a fixed stellar rank reveals that code structure adapts to specific noise types. Grid-like arrangements of quantum information prove beneficial when photons are lost, while rotationally symmetric designs better withstand dephasing, a process where quantum information degrades due to environmental interactions. Achieving break-even performance, where error correction is effective, with a stellar rank of two demonstrates a significant reduction in required resources for stabilising quantum information against dephasing.

Researchers demonstrated that optimising bosonic quantum error correction codes using stellar rank, a measure of state complexity, reveals a trade-off between code quality, energy and protection against noise. This is important because it establishes a quantifiable link between the resources needed to prepare quantum codes and their ability to correct errors caused by photon loss and dephasing. The study found that a stellar rank of two is sufficient to achieve effective error correction against dephasing, and that code structure adapts to the type of noise present. The authors suggest further work is needed to determine if these findings apply to a wider range of quantum encoding methods.