本文關(guān)鍵詞： LDPC碼 解碼器 誤差平底 解碼算法 乘積碼 polar碼　出處：《南京大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

【摘要】：隨著現(xiàn)代數(shù)字通信的進一步發(fā)展,LDPC碼憑借其優(yōu)異的糾錯性能被越來越多的通信標準所采用。2006年由國家廣電總局頒布的中國移動多媒體廣播標準(CMMB)就采用了LDPC碼。本文基于Layered min-sum解碼算法,為了降低解碼器功耗,提出了一種非規(guī)則量化策略,并設(shè)計了相應(yīng)的應(yīng)用于CMMB標準的解碼器及各模塊結(jié)構(gòu)。仿真結(jié)果表明,該設(shè)計方案能夠在保證必要吞吐率的前提下減少約35%的計算,從而降低了解碼器功耗。 目前,LDPC碼的研究領(lǐng)域主要圍繞著兩個大的方面：一是解碼器的設(shè)計；二是碼的性能,尤其是在高信噪比區(qū)域,對誤差平底的研究。為了克服誤差平底,本文針對LDPC-SPC乘積碼,提出了一種相應(yīng)的行列混合迭代解碼算法。實驗結(jié)果表明,該乘積碼理論上能夠有效降低誤差平底,且解碼復(fù)雜度僅稍有增加,對降低LDPC碼誤差平底的研究提供了一種可選方案。 Polar碼即極化碼,是E.Arikan在2007年提出的一種最新的編碼技術(shù)。在二進制離散無記憶信道條件下,極化碼理論上被證明可以達到香農(nóng)極限。近來,極化碼因其接近香農(nóng)限的性能,且編解碼具有較低復(fù)雜度,逐漸成為糾錯碼研究的熱點。本文提出了一種簡單的基于部分并行輸入的極化碼編碼器的硬件結(jié)構(gòu)實現(xiàn),有效解決了極化碼由于碼字過長而編碼復(fù)雜度較高的問題。 最后,本文討論了極化碼的譯碼算法。由于傳統(tǒng)的連續(xù)消除譯碼算法是串行解碼,所以實際應(yīng)用中所需要的長碼字會造成很高的解碼時延。本文提出了一種改進的連續(xù)消除譯碼算法,這種改進后的譯碼算法相比于傳統(tǒng)的譯碼算法能減少約一半的譯碼延遲。而且,仿真結(jié)果表明改進后的解碼算法在糾錯性能上的損失是可以忽略不計的。
[Abstract]:With the further development of modern digital communication. LDPC code is adopted by more and more communication standards because of its excellent error-correcting performance. In 2006, China Mobile Multimedia Broadcasting Standard was issued by State Administration of Radio, Film and Television (SARFT). LDPC code is adopted. This paper is based on Layered min-sum decoding algorithm. In order to reduce the power consumption of the decoder, an irregularity quantization strategy is proposed, and the corresponding decoder and each module structure applied to the CMMB standard are designed. The simulation results show that. The proposed scheme can reduce the power consumption of the decoder by about 35% while ensuring the necessary throughput. At present, the research field of LDPC codes mainly revolves around two aspects: first, the design of decoder; The second is the performance of the code, especially in the high SNR region, the study of the error flat bottom. In order to overcome the error flat bottom, this paper aims at the LDPC-SPC product code. An iterative decoding algorithm for column and column mixing is proposed. Experimental results show that the product code can effectively reduce the error at flat bottom theoretically, and the decoding complexity is only slightly increased. An alternative scheme is provided for the research of reducing the error of LDPC codes at flat bottom. Polar code is called polarization code, which is the latest coding technique proposed by E. Arikan in 2007. In the condition of binary discrete memoryless channel. Polarization codes have been proved to reach Shannon limit theoretically. Recently, polarization codes have low complexity due to their performance close to Shannon limit. This paper presents a simple hardware implementation of polarization code encoder based on partial parallel input. The problem of high coding complexity due to long codewords is effectively solved. Finally, this paper discusses the decoding algorithm of polarimetric code, because the traditional continuous cancellation decoding algorithm is serial decoding. Therefore, the long codewords needed in practical applications will cause a high decoding delay. In this paper, an improved continuous decoding algorithm is proposed. Compared with the traditional decoding algorithm, the improved decoding algorithm can reduce the decoding delay by about half, and the simulation results show that the loss of the improved decoding algorithm in error correction performance is negligible.
【學(xué)位授予單位】：南京大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：TN911.22;TN915.1

【參考文獻】

相關(guān)期刊論文 前1條

1 喬華;林廣榮;王達;董明科;項海格;;LDPC-SPC乘積碼[J];北京大學(xué)學(xué)報(自然科學(xué)版);2008年01期

，

本文編號：1467476