Mitigating measurement readout errors in NISQ devices

Abstract

The currently available quantum technologies called noisy intermediate-scale quantum (NISQ) technologies containing noise in all of the steps, preparation, dynamics, and measurement do not yet suffice to show the quantum advantages in general although quantum computation in the ideal manner can outperform the classical counterparts. One of the key challenges, while quantum error correction is not feasible either, is therefore to reduce the effects of noise and run quantum algorithm on NISQ devices. In this thesis, we introduce a method of mitigating measurement readout errors on realistic quantum detectors and consider quantum algorithms on NISQ devices. The method applies i) single-qubit gate based quantum pre-processing before a measurement and then ii) classical post-processing after a measurement. We show that quantum algorithms such as the Bernstein-Vazirani and amplitude amplification algorithms on NISQ devices can be improved by the proposed measurement-mitigation protocol. Proof-of-principle demonstrations are also shown in IBMQ_yorktown and IBMQ_essex.