Skip to main navigation Skip to search Skip to main content

Quantum multi-anomaly detection

Santiago Llorens, Gael Sentís, Ramon Muñoz-Tapia

Research output: Contribution to journalArticleResearchpeer-review

Abstract

A source assumed to prepare a specified reference state sometimes prepares an anomalous one. We address the task of identifying these anomalous states in a series of n preparations with k anomalies. We analyze the minimum-error protocol and the zero-error (unambiguous) protocol and obtain closed expressions for the success probability when both reference and anomalous states are known to the observer and anomalies can appear equally likely in any position of the preparation series. We find the solution using results from association schemes theory, thus establishing a connection between graph theory and quantum hypothesis testing. In particular, we use the Johnson association scheme which arises naturally from the Gram matrix of this problem. We also study the regime of large n and obtain the expression of the success probability that is non-vanishing. Finally, we address the case in which the observer is blind to the reference and the anomalous states. This scenario requires a universal protocol for which we prove that in the asymptotic limit, the success probability corresponds to the average of the known state scenario.

Original languageEnglish
Pages (from-to)1452-1452
JournalQuantum
Volume8
DOIs
Publication statusPublished - 28 Aug 2024

Fingerprint

Dive into the research topics of 'Quantum multi-anomaly detection'. Together they form a unique fingerprint.

Cite this