A High-Speed Successive-Cancellation Decoder for Polar Codes Using Approximate Computing
Yangcan Zhou, Zhiyu Chen, Jun Ting Lin, Zhongfeng Wang · IEEE Transactions on Circuits & Systems II Express Briefs · 2018
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.