Qubit fidelity under stochastic Schr\"odinger equations driven by colored noise

TL;DR

This paper develops a method to analyze qubit fidelity under realistic colored noise, such as Ornstein-Uhlenbeck noise, providing detailed statistical insights beyond traditional Lindblad models.

Contribution

It introduces a novel approach for solving the full distribution of qubit fidelity under realistic stochastic Schrödinger equations, extending beyond white noise assumptions.

Findings

Allows prediction of mean, variance, and higher moments of fidelity.

Enables assessment of noise levels for quantum system quality.

Facilitates optimal control strategies under classical noise.

Abstract

Environmental noise on a controlled quantum system is generally modeled by a dissipative Lindblad equation. This equation describes the average state of the system via the density matrix $\rho$. One way of deriving this Lindblad equation is by introducing a stochastic operator evolving under white noise in the Schr\"odinger equation. However, white noise, where all noise frequencies contribute equally in the power spectral density, is not a realistic noise profile as lower frequencies generally dominate the spectrum. Furthermore, the Lindblad equation does not fully describe the system as a density matrix $\rho$ does not uniquely describe a probabilistic ensemble of pure states $\{\psi_j\}_j$. In this work, we introduce a method for solving for the full distribution of qubit fidelity driven by important stochastic Schr\"odinger equation cases, where qubits evolve under more realistic noise profiles, e.g. Ornstein-Uhlenbeck noise. This allows for predictions of the mean, variance, and higher-order moments of the fidelities of these qubits, which can be of value when deciding on the allowed noise levels for future quantum computing systems, e.g. deciding what quality of control systems to procure. Furthermore, these methods will prove to be integral in the optimal control of qubit states under (classical) control system noise.