标题： 量子拓扑码的气泡聚类解码器 显示英文标题

标题： Bubble Clustering Decoder for Quantum Topological Codes

摘要： 量子计算机对噪声非常敏感，需要使用纠错码来保护存储的数据。 必须持续校正错误以对抗退相干，使用适当的解码器。 因此，能够处理实时奇偶校验提取的快速解码策略对于实现容错量子计算至关重要。 在本文中，我们介绍了用于量子表面码的气泡聚类（BC）解码器，它作为MWPM的低延迟替代方案，在牺牲轻微性能下降的情况下实现了显著更快的执行速度。 这种速度提升是通过基于缺陷中心的气泡进行高效的聚类生成获得的，并避免了传统解码器中常用的聚类增长和合并阶段所带来的计算开销。 我们的复杂度分析表明，所提出的解码器的复杂度与缺陷数量的平方成比例。 对于中等物理错误率，这相当于数据量子比特数量的线性复杂度。 显示英文摘要

摘要： Quantum computers are highly vulnerable to noise, necessitating the use of error-correcting codes to protect stored data. Errors must be continuously corrected over time to counteract decoherence using appropriate decoders. Therefore, fast decoding strategies capable of handling real-time syndrome extraction are crucial for achieving fault-tolerant quantum computing. In this paper, we introduce the bubble clustering (BC) decoder for quantum surface codes, which serves as a low-latency replacement for MWPM, achieving significantly faster execution at the cost of a slight performance degradation. This speed boost is obtained leveraging an efficient cluster generation based on bubbles centered on defects, and avoiding the computational overhead associated with cluster growth and merging phases, commonly adopted in traditional decoders. Our complexity analysis reveals that the proposed decoder operates with a complexity on the order of the square of the number of defects. For moderate physical error rates, this is equivalent to linear complexity in the number of data qubits.