Sampling Noise and Optimized Measurement Distribution in Imaginary-Time Quantum Dynamics Simulations

TL;DR

This paper analyzes how sampling noise affects variational quantum dynamics simulations and proposes optimized measurement strategies to improve ground-state preparation accuracy on noisy quantum devices.

Contribution

It introduces measurement-distribution optimization methods and demonstrates their effectiveness in reducing measurement costs and improving fidelity in noisy imaginary-time quantum simulations.

Findings

Optimized shot allocation significantly improves state fidelity.

Tikhonov regularization stabilizes imaginary-time evolution under noise.

Distributing shots evenly among circuits yields the best results.

Abstract

Variational quantum dynamics simulations (VQDS) provide a promising route to simulate real- and imaginary-time quantum dynamics on noisy intermediate-scale quantum devices using fixed-depth circuits. However, their practical performance is strongly limited by sampling noise arising from a finite number of circuit measurements. In this work, we systematically investigate the impact of sampling noise on VQDS, with a focus on ground-state preparation in one-dimensional Ising spin models using imaginary time evolution. We compare different regularization strategies for stabilizing the equations of motion and show that Tikhonov regularization provides robust performance in noisy imaginary-time evolution. We then benchmark measurement-distribution strategies that allocate shots by minimizing a cost function that characterizes the error in solving the equation of motion. Using noisy circuit simulations, we demonstrate that such optimized shot allocation can significantly improve state fidelity and reduce the total measurement cost by more than a factor of two compared to uniform shot distributions. We observe that the best results are found if a sufficiently large number of measurements is guaranteed for all circuits, suggesting that a finite fraction of shots should be distributed evenly. Our results provide practical guidelines for implementing measurement-efficient variational quantum dynamics and ground-state preparation on near-term quantum hardware.