IBM-Q 超導量子電腦之量子閘雜訊特徵描述與分析研究

Abstract

目前的量子電腦被稱為雜訊中等規模量子電腦（NISQ電腦），發展受到雜訊影響而受限，只能進行小規模計算且無法達成預期的應用。為了實現量子電腦的實際應用，了解並分析雜訊有助於保護量子資訊受到影響和 NISQ 電腦的開發。目前測量錯誤率的主要方法是隨機基準測試（RB）。 RB方法將來自不同量子邏輯閘的雜訊視為相同的雜訊，然而若不同的量子邏輯閘具有不同類型的雜訊，這種測量錯誤率方法就會有很多的討論空間。本研究介紹了一種定量特徵化量子邏輯閘雜訊的方法，系統性地分析了 IBM-Q 量子系統中的雜訊特性。透過量子動力學的擬合方法，我們得出了有關 IBM-Q 裝置雜訊特性的寶貴見解。隨後，我們採用統計方法建立了一個物理模型，深入分析了這些雜訊來源。利用開放量子理論框架，我們成功地藉由量子主方程式模擬了耗散系統的動態。對量子電腦進行詳細的雜訊特徵化除了有助於模擬開放量子系統動態中的量子雜訊之外，對於未來應用在量子電腦上的錯誤修正演算法或錯誤緩解演算法至關重要，有助於現階段量子電腦的發展。

Noisy Intermediate-Scale Quantum (NISQ) devices define the current state of quantum computing, capable of executing modest-scale calculations. Analyzing noise is crucial for protecting quantum information from errors and advancing these devices. This research introduces a method for quantitatively characterizing gate-specific noise within IBM-Q systems. Through a fitting procedure based on quantum dynamics and employing stochastic methodologies, we analyze noise profiles using the stochastic Liouville equation. This detailed characterization aids in simulating quantum noises in open quantum systems, a significant application of NISQ devices. We decompose noise using known channels and the quantum master equation, determining error rates and the noise Hamiltonian, and reproducing quantum dynamics to reveal environmental noise impacts. Observations of non-Gaussian distributions and correlated noises have significant implications for noise mitigation and error correction. Our findings suggest that specific gate sequences can be designed to mitigate or transform certain types of quantum noise, enhancing quantum simulation, error mitigation, and Quantum Error Correction (QEC). Understanding these noise mechanisms in IBM-Q computers is crucial for developing effective error mitigation protocols and shaping the future of quantum computing.