Sliding window decoding has been employed as the decoding scheme of quasi-cyclic spatially-coupled low-density parity-check (QC-SC-LDPC) codes, with the decoding latency and complexity independent of the coupling length but proportional to the window size. The window size was empirically chosen at least three times the constraint length in previous works, which results in the preference of QC-SC-LDPC codes with small constraint length, to reduce the decoding latency and complexity in practical systems. However, the code design freedom and the decoder throughput are limited for these QC-SC-LDPC codes. In this paper, the optimal window size under the constraint of practical decoding complexity is analyzed employing multi-edge-type density evolution (MET-DE). It can be verified from the MET-DE analytical result that the optimal window size could achieve less than twice the constraint length for several edge spreading patterns, which largely loose this restriction in practical code design. Furthermore, it could be observed that besides the constraint length, the specific structure of the edge spreading pattern could affect the optimal window size as well, and thus the MET-DE analytical method for the optimal window size could effectively guide the design of QC-SC-LDPC codes in practical systems. Then an improved sliding window decoding scheme is proposed, including the optimal choice of window size and a parity-check based early termination scheme. Finally, the performance and robustness of the proposed sliding window decoding scheme for QC-SC-LDPC codes are demonstrated via simulation, with the decoding complexity at most twice as that of the conventional LDPC coded schemes.

中文翻译：

---

实现复杂度约束下的QC-SC-LDPC码滑窗解码

滑动窗口解码已被用作准循环空间耦合低密度奇偶校验 (QC-SC-LDPC) 码的解码方案，其解码延迟和复杂度与耦合长度无关，但与窗口大小成正比。在以前的工作中，窗口大小根据经验选择至少是约束长度的三倍，这导致了约束长度较小的 QC-SC-LDPC 码的偏好，以减少实际系统中的解码延迟和复杂性。然而，这些 QC-SC-LDPC 码的代码设计自由度和解码器吞吐量是有限的。在本文中，采用多边型密度演化（MET-DE）分析了实际解码复杂度约束下的最佳窗口大小。从 MET-DE 分析结果可以验证，对于几种边缘扩展模式，最佳窗口大小可以实现小于两倍的约束长度，这在实际代码设计中大大放松了这一限制。此外，可以观察到除了约束长度之外，边缘扩展模式的特定结构也会影响最佳窗口大小，因此最佳窗口大小的 MET-DE 分析方法可以有效指导 QC-实际系统中的 SC-LDPC 码。然后提出了一种改进的滑动窗口解码方案，包括窗口大小的最优选择和基于奇偶校验的提前终止方案。最后，通过仿真证明了所提出的 QC-SC-LDPC 码滑动窗口解码方案的性能和鲁棒性，