Comparing quantum incompatibility of device sets from an operational perspective
Kensei Torii, Ryo Takakura, Ryotaro Imamura
TL;DR
The paper introduces an operational preorder $\preceq_{\mathrm{inc}}$ to compare quantum device incompatibility based on how readily incompatibility is detected with restricted state sets $\mathcal{S}_0$. This ordering preserves fundamental monotonicity under convex mixing with compatible devices, observable post-processing, and channel concatenation, and it yields a fine-grained hierarchy beyond traditional measures. In the prominent case of unbiased qubit observables, the induced equivalence $\sim_{\mathrm{inc}}$ uniquely identifies mutually unbiased qubit pairs, contrasting them with all other unbiased pairs, and links the ordering to the distributed sampling task as a QC certification tool. Numerically, the authors map regions where certain Mub configurations are ordered, revealing new classifications and suggesting that a restricted family of state sets $\mathcal{S}_0^{(R)}$ suffices to characterize the ordering for these observables. The work paves the way for applying the ordering to broader quantum-device scenarios and GPT frameworks, with potential extensions and computational approaches (e.g., SDP) for higher-dimensional systems.
Abstract
To effectively utilize quantum incompatibility as a resource in quantum information processing, it is crucial to evaluate how incompatible a set of devices is. In this study, we propose an ordering to compare incompatibility and reveal its various properties based on the operational intuition that larger incompatibility can be detected with fewer states. We especially focus on typical class of incompatibility exhibited by mutually unbiased qubit observables and numerically demonstrate that the ordering yields new classifications among sets of devices. Moreover, the equivalence relation induced by this ordering is proved to uniquely characterize mutually unbiased qubit observables among all pairs of unbiased qubit observables. The operational ordering also has a direct implication for a specific protocol called distributed sampling.
