Quaternary-Binary Message-Passing Decoder for Quantum LDPC Codes

Quaternary-Binary Message-Passing Decoder for Quantum LDPC Codes
复制标题

DOI:
10.1109/globecom54140.2023.10436874
复制
发表时间:
2023-12
期刊:
GLOBECOM 2023 - 2023 IEEE Global Communications Conference
影响因子:
--
通讯作者:
Dimitris Chytas;Nithin Raveendran;Asit Kumar Pradhan;Bane V. Vasic
Dimitris Chytas;Nithin Raveendran;Asit Kumar Pradhan;Bane V. Vasic
中科院分区:
其他
文献类型:
--
作者:
Dimitris Chytas;Nithin Raveendran;Asit Kumar Pradhan;Bane V. Vasic

文献摘要

相似文献

提出了一种低复杂度的量子低密度奇偶校验(QLDPC)稳定码的消息传递量子纠错算法。建议的解码器上操作的四元稳定图,但只交换二进制消息。这导致了显着降低复杂性相比,其他四元信念传播(BP)算法,通过浮点消息。建议的解码器的效率进行评估,通过提供解码的例子,性能指标,使用蒙特-卡罗模拟,和复杂性分析。尽管其降低了复杂性,所提出的解码器的性能损失是适度的相比,浮点并行四进制解码器的Calderbank-Shor-Steane(CSS)代码的家庭。特别是,在[[1054,140,20]]提升的产品(LP)坦纳码上获得的实验表明,对于低错误率(< 0.01),所提出的四进制-二进制消息传递解码器通过在几乎相同的迭代次数中收敛而接近四进制BP的性能,同时需要更少的复杂操作。此外,对于非CSS代码,我们的解码器执行类似的四进制浮点解码器,尽管其较低的复杂度。
We introduce a low-complexity message-passing quantum error correction algorithm for decoding Quantum Low-Density Parity-Check (QLDPC) stabilizer codes. The proposed decoder operates on the quaternary stabilizer graph but only exchanges binary messages. This leads to a significantly reduced complexity compared to other quaternary belief propagation (BP) algorithms that pass floating-point messages. The efficacy of the proposed decoder is evaluated by providing decoding examples, performance metrics using Monte-Carlo simulations, and complexity analysis. Despite its reduced complexity, the performance loss of the proposed decoder is modest compared to floating-point parallel quaternary decoders for a Calderbank-Shor-Steane (CSS) code family. In particular, experiments obtained over the [[1054, 140, 20]] lifted product (LP) Tanner code demonstrated that for low error rates (< 0.01), the proposed quaternary-binary message-passing decoder approaches the performance of quaternary BP by converging in almost the same number of iterations while requiring less complex operations. Additionally, for non-CSS codes, our decoder performs similarly as quaternary floating-point decoders despite its lower complexity.