【摘要】 准循环低密度奇偶校验(quasi-cyclic low-density parity-check,QC-LDPC)码是一种应用广泛的编码技术,该技术主体包含校验部分和信息部分。现有的编码技术主要针对校验部分进行研究改进,而缺乏对信息矩阵的构造来提升编码性能,并且信息部分和校验部分相互独立从而降低了编码的性能。针对该问题,提出一种大围长可快速编码的QC-LDPC码构造方法,该方法将最大公约数(greatest common divisor,GCD)算法、行列加减值和掩饰技术引入到校验矩阵得到一种改进型下三角结构的校验矩阵,构造出的QC-LDPC码兼容了大围长和低编码复杂度的双重特性,从而提升编码灵活性。仿真结果显示与GCD算法构造的围长为8的QC-LDPC码相比较,提出的快速编码方法在误码率(bit error rate,BER)为10-5时获得0.25 d B的编码增益;与基于渐进边长(progress edge growth,PEG)算法构造的随机码相比较,构造的非规则QC-LDPC码在误码率为10-5时码字性能提高了约0.1 d B。更多还原

【Abstract】 The quasi-cyclic low-density parity-check codes( QC-LDPC) is a widely used coding technique. It’s main body includes check and information. The existing coding technology mainly focuses on the check part,but lacks the construction of the information matrix to improve the coding performance. And the information part and check part are independent of each other,thus reducing the performance of coding. Aiming at this problem,a method for constructing QC-LDPC codes with large girth and fast encoding is proposed. The method adopts the greatest common divisor( GCD) algorithm,row-column plus and minus values and masking technology to get a check matrix for the improved lower triangular structure,which makes the constructed QC-LDPC codes compatible with the double peculiarities of large girth and low encoding complexity,and the encoding flexibility is higher. The simulation results show that compared with the GCD algorithm with girth 8 QCLDPC code,the proposed GCD algorithm for fast encoding method based on bit error rate( BER) of 10-5 is obtained by 0.25 d B encoding gain and compared with the random codes constructed by progress edge growth( PEG) algorithm,the constructed irregular QC-LDPC codes improve the code character performance about 0.1 d B at bit error rate of 10-5.更多还原