Open Access
ARTICLE
LIRB-Based Quantum Circuit Fidelity Assessment and Gate Fault Diagnosis
Laboratory for Advanced Computing and Intelligence Engineering, Information Engineering University, Zhengzhou, 450001, China
* Corresponding Author: Zheng Shan. Email:
Computers, Materials & Continua 2025, 82(2), 2215-2233. https://doi.org/10.32604/cmc.2024.058163
Received 05 September 2024; Accepted 14 November 2024; Issue published 17 February 2025
Abstract
Quantum circuit fidelity is a crucial metric for assessing the accuracy of quantum computation results and indicating the precision of quantum algorithm execution. The primary methods for assessing quantum circuit fidelity include direct fidelity estimation and mirror circuit fidelity estimation. The former is challenging to implement in practice, while the latter requires substantial classical computational resources and numerous experimental runs. In this paper, we propose a fidelity estimation method based on Layer Interleaved Randomized Benchmarking, which decomposes a complex quantum circuit into multiple sublayers. By independently evaluating the fidelity of each layer, one can comprehensively assess the performance of the entire quantum circuit. This layered evaluation strategy not only enhances accuracy but also effectively identifies and analyzes errors in specific quantum gates or qubits through independent layer evaluation. Simulation results demonstrate that the proposed method improves circuit fidelity by an average of 6.8% and 4.1% compared to Layer Randomized Benchmarking and Interleaved Randomized Benchmarking methods in a thermal relaxation noise environment, and by 40% compared to Layer RB in a bit-flip noise environment. Moreover, the method detects preset faulty quantum gates in circuits generated by the Munich Quantum Toolkit Benchmark, verifying the model’s validity and providing a new tool for faulty gate detection in quantum circuits.Keywords
Cite This Article

This work is licensed under a Creative Commons Attribution 4.0 International License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.