第五代行動通訊之低密度校驗碼解碼器於繪圖處理器平行化實作

GPU-based Parallel Decoder of Low-Density Parity-Check Code for 5G New Radio

梁育齊(Yu-Chi Liang)

指導教授 ： 許騰尹

國立交通大學/資訊學院/資訊科學與工程研究所/碩士(2020年)

摘要

在機器運算能力快速增長的環境下，從六零年代就已經被提出的低密度校驗碼，又再一次的回到了通訊編碼演算法的討論行列中，並且在第五代行動通訊網路(5th Generation New Radio)規格裡，被定為在資料傳輸通道中的編碼演算法；然而，其解碼演算法中的運算量非常大，導致了很長的解碼延遲。本文分析了運算時資料的相依性，並提出平行運算的架構來提高運算的效率，主要使用適合做大量平行運算的繪圖處理器(Graphic Processing Unit, GPU)，搭配CUDA(Compute Unified Device Architecture)介面，並且提出適合演算法的平行優化策略，來增進解碼演算法的運算效率，在單編碼區塊的解碼時間較CPU快了260%。本文將著重於平行優化架構的編排，並探討和分析其執行效率以及資源使用情況。

關鍵字

繪圖處理器 ； 第五代行動通訊網路 ； 通道解碼 ； 軟體基地台 ； 實體層

並列摘要

As the computation ability rapidly increases, Low-Density Parity-Check Code is once again back to the discussing list of channel coding algorithm after it was first invented in 1960th. Its outstanding error recover ability earns itself a place in the specification of the 5th Generation Mobile Networks (5G New Radio) as the encoding algorithm for data channel. However, the counterpart, decoding algorithm, possesses a great amount of arithmetic calculation and tremendously high complexity. This thesis first evaluates the data processed in the decoding procedure, then proposes several parallel execution flows to increase the decoding efficiency. In single code block, the decoding latency is significantly improved by 260%. This thesis utilizes GPU (graphic processing unit) with CUDA frame work to achieve parallel calculation, proposes several parallel execution architectures suitable for LDPC decoding process, and analyses the efficiency and GPU resource utilization.

並列關鍵字

Graphic Processing Unit, GPU ； 5G New Radio ； LDPC ； Channel Coding ； Physical Layer

參考文獻

[1] R. Gallager, “Low-density parity-check codes,” IRE Trans. Inf. Theory, vol. 8, no. 1, pp. 21–28, Jan. 1962.

[2] R. M. Tanner, “A recursive approach to low complexity codes,” IEEE Trans. Inf. Theory, vol. 27, no. 5, pp. 533–547, Sep. 1981.

[3] D. J. C. MacKay and R. M. Neal, “Near Shannon limit performance of low density parity check codes,” Electron. Lett., vol. 32, no. 18, pp. 1645–1646, Aug. 1996.

[4] J. Chen and M. P. C. Fossorier, “Density evolution for two improved BP-based decoding algorithms of LDPC codes,” IEEE Commun. Lett., vol. 6, no. 5, pp. 208–210, May 2002.

[5] X.-Y. Hu, E. Eleftheriou, and D. M. Arnold, “Progressive edge-growth Tanner graphs,” in Proc. IEEE Global Telecommun. Conf. (GlobeCom), San Antonio, TX, USA, Nov. 2001, pp. 995–1001.