Self-guided tomography of time-frequency qudits

TL;DR

This paper demonstrates a scalable, high-fidelity method for characterizing high-dimensional time-frequency quantum states using self-guided tomography, which is robust against noise and does not require calibration.

Contribution

It extends self-guided tomography to the time-frequency domain, enabling precise, calibration-free state estimation for high-dimensional quantum systems.

Findings

Achieved >99% fidelity for 3- and 5-dimensional states

Demonstrated robustness against noise and environmental disturbances

Validated scalability for larger Hilbert spaces

Abstract

High-dimensional time-frequency encodings have the potential to significantly advance quantum information science; however, practical applications require precise knowledge of the encoded quantum states, which becomes increasingly challenging for larger Hilbert spaces. Self-guided tomography (SGT) has emerged as a practical and scalable technique for this purpose in the spatial domain. Here, we apply SGT to estimate time-frequency states using a multi-output quantum pulse gate. We achieve fidelities of more than 99\% for 3- and 5-dimensional states without the need for calibration or post-processing. We demonstrate the robustness of SGT against statistical and environmental noise, highlighting its efficacy in the photon-starved regime typical of quantum information applications.