Minimising statistical errors in calibration of quantum-gate sets

TL;DR

This paper enhances quantum gate calibration by analyzing measurement uncertainties, optimizing measurement settings, and demonstrating that adding two single-qubit gates significantly reduces statistical errors in calibrating multi-qubit gates.

Contribution

It extends the GSC protocol with a statistical analysis and optimized measurement settings, improving calibration accuracy with minimal additional gates.

Findings

Measurement uncertainties are quantitatively analyzed.

Optimized measurement settings reduce statistical errors.

Adding two single-qubit gates halves the calibration error for CNOT gates.

Abstract

Calibration of quantum gates is a necessary hurdle to overcome on the way to a reliable quantum computer. In a recent paper, a protocol called Gate Set Calibration protocol (GSC) has been introduced and used to learn coherent errors from multi-qubit quantum gates. Here, we extend this study in a number of ways: First, we perform a statistical analysis of the measurement uncertainties. Second, we find explicit measurement settings that minimize this uncertainty, while also requiring that the protocol involves only a small number of distinct gates, aiding physical realizability. We numerically demonstrate that, just by adding two more single-qubit gates to GSC, the statistical error produced in the calibration of a CNOT gate is divided by a factor of more than two.