Easy better quantum process tomography

TL;DR

This paper introduces a method to improve quantum process tomography by incorporating SPAM calibration data, reducing bias and gauge freedom, and providing explicit correction procedures for linear-inversion techniques.

Contribution

It presents a novel correction procedure for standard and overcomplete linear-inversion QPT that leverages SPAM calibration data to enhance accuracy.

Findings

Reduces bias in quantum process estimates

Provides explicit correction procedures for linear-inversion QPT

Extends correction methods to maximum likelihood estimators

Abstract

Quantum process tomography (QPT), used to estimate the linear map that best describes a quantum operation, is usually performed using a priori assumptions about state preparation and measurement (SPAM), which yield a biased and inconsistent estimator. This estimate can be made more accurate and less biased by incorporating SPAM-calibration data. Unobservable properties of the SPAM operations introduce a small gauge freedom that can be regularized using the a priori SPAM. We give an explicit correction procedure for standard linear-inversion QPT and overcomplete linear-inversion QPT, and describe how to extend it to statistically principled estimators like maximum likelihood estimation.