A High-Speed Successive-Cancellation Decoder for Polar Codes Using Approximate Computing
A High-Speed Successive-Cancellation Decoder for Polar Codes Using Approximate Computing
复制标题
DOI：
10.1109/tcsii.2018.2847441
复制
发表时间：
2019-02
期刊：
影响因子：
--
通讯作者：
Yangcan Zhou;Zhiyu Chen;Jun Lin;Zhongfeng Wang
中科院分区：
文献类型：
--
作者：
Yangcan Zhou;Zhiyu Chen;Jun Lin;Zhongfeng Wang
极化码是一类新的前向纠错码，已被证明在对称无记忆信道下具有渐近容量实现。连续消除（SC）解码器复杂度低，同时由于其串行处理性质而遭受低速度。基于一个新公布的快速简化SC算法，我们提出了一个更有效的解码器比现有技术在这个简短的。在解码器核心的组件进行了优化开发。此外，几个有效的近似计算单元（Apx-CU）的引入，以取代他们的准确对应的解码器，以实现更高的工作频率。现场可编程门阵列的实现结果表明，我们的设计是约1.5倍，比现有技术，和建议的Apx-CU可以带来额外的19%的处理速度。
Polar codes are a new class of forward-error-correction codes, which have been proved asymptotically capacity-achieving for symmetric memoryless channels. Successive-cancellation (SC) decoders are low complexity while suffering from low speed due to their serial processing nature. Based on a newly published fast simplified SC algorithm, we propose a more efficient decoder than the prior art in this brief. Well-optimized components in the decoder core are developed. Additionally, several efficient approximate computing units (Apx-CUs) are introduced to substitute their accurate counterparts in the decoder to achieve even higher operating frequency. Field-programmable gate-array implementation results show that our design is about 1.5 times faster than the prior art, and the proposed Apx-CUs can bring an additional 19% processing speed.