Real-time diagnostics on a QKD link via QBER Time Series Analysis

TL;DR

This paper presents a real-time machine learning-based method for diagnosing impairments in quantum key distribution links by analyzing QBER and SKR time series, enhancing operational awareness.

Contribution

It introduces a supervised ML pipeline that identifies impairments in QKD channels in real time without system intervention, addressing current technological limitations.

Findings

Effective real-time impairment detection using QBER and SKR data

No additional interventions needed for diagnosis

Potential to improve QKD network reliability

Abstract

The integration of QKD systems in Metro optical networks raises challenges which cannot be completely resolved with the current technological status. In this work we devise a methodology for identifying different kind of impairments which may occur on the quantum channel during its transmission in an operational network. The methodology is built around a supervised ML pipeline which is using as input QBER and SKR time-series and requires no further interventions on the QKD system. The identification of impairments happens in real time and even though such information cannot reverse incidents, this can be valuable for users, operators and key management system.