Enhanced Algorithmic Perfect State Transfer on IBM Quantum Computers

TL;DR

This paper investigates the implementation of perfect state transfer on IBM quantum computers, models noise effects comprehensively, and proposes mitigation techniques to improve transfer success probability.

Contribution

It introduces a detailed noise model for quantum state transfer and demonstrates noise mitigation strategies, advancing practical quantum communication protocols.

Findings

Simulations closely match experimental results with a comprehensive noise model.

Noise mitigation techniques improve success probability by over 27% in simulators.

Optimized couplings enhance transfer success probability, with improvements up to 26% in simulations.

Abstract

Perfect state transfer (PST) through a spin chain can be theoretically obtained via predesigned PST couplings. However, the corresponding experiment on IBM quantum computers demonstrates low transmission success probability (SP) due to noises. Using few qubits of their 127-qubit Eagle processors, we perform the simulation of algorithmic PST through an XY spin chain with PST couplings on ibm_sherbrooke and ibm_brisbane processors, alongside Qiskit simulations. The peak SP cannot reach 1 ($\sim$0.725 peak SP for N=4). We then propose a comprehensive noise model including Pauli errors, thermal relaxation ($T_1$) and dephasing ($T_2$), and ZZ crosstalk. Based on the experimental parameters provided by the IBM superconducting quantum computing platform, we perform the Qiskit simulation with the comprehensive noise model, and find that the time evolution of the SP is highly consistent with the experimental results. This simulation yields a peak SP of 0.761 at $\textstyle t\approx\pi/4$, closely matching the results on hardware. To mitigate the impact of noise, we use rescaling techniques to correct noise-induced time shifts and SP decay, achieving an SP improvement of 0.210 (27.60%) in simulators and 0.263 (38.23%) on hardware, aligning hitting times closer to ideal values. Additionally, optimal couplings designed via grid search and refined by Bayesian optimization under the comprehensive noise model achieve an SP improvement of 0.190 (26.21%) in simulators and 0.056 (7.72%) on hardware. Our work highlights challenges in implementing algorithmic PST on current quantum computers, proposes a comprehensive noise model to effectively describe the system dynamics, and provides insights for developing noise-robust quantum communication protocols.