Quantum Analog-to-Digital Converter for Phase Estimation

TL;DR

This paper demonstrates the first experimental quantum analog-to-digital converter for phase estimation, surpassing classical limits using entangled states and machine learning for noise reduction.

Contribution

It introduces an experimental quantum digital estimation protocol that converts continuous parameters into bit strings, surpassing the standard quantum limit.

Findings

Achieved digital phase estimation exceeding the standard quantum limit.

Implemented noise deconvolution using machine learning techniques.

Benchmarking shows improved information recovery over classical strategies.

Abstract

Traditional quantum metrology assesses precision using the figures of merit of continuous-valued parameter estimation. Recently, quantum digital estimation was introduced: it evaluates the performance information-theoretically by quantifying the number of significant bits of the parameter, redefining key benchmarks like the Heisenberg bound. Here, we report the first experimental realization of a Quantum Analog-to-Digital Converter for quantum metrology, that takes an input continuous parameter and outputs a bit string, using an advanced photonic platform, comprising a fully reconfigurable integrated circuit and a quantum dot source of highly indistinguishable photons. We implement a protocol for digital phase estimation that is capable of surpassing the standard quantum limit, through the simultaneous use of different entangled state resources. We tackle experimental imperfections by employing machine learning techniques for noise deconvolution and estimation process refinement. Our protocol is experimentally benchmarked against classical strategies via the number of recoverable bits of the unknown parameter. Our results open new perspectives for future implementation of quantum digital estimation strategies.