优化稀疏LU分解的LDPC编码算法研究

中国移动多媒体广播标准中采用了高度结构化低密度校验码编码.针对该低密度校验码非系统码和非准循环特性,设计了一种基于LU分解的编码算法,巧妙地设计了紧凑的矩阵存储方式,修正了行重列重乘积最小化的优化准则,提出了优化的稀疏LU分解.仿真计算结果表明,相对其他文献中的算法,所提稀疏LU分解得到的L和U中元素"1"的个数减少10%,具有更好的稀疏性,进一步降低了编码算法的时间复杂度.因此,基于优化稀疏LU分解的低密度校验码编码算法在中国移动多媒体广播标准实际应用中具有较好的前景.

LDPC encoding algorithm based on optimized sparse LU decomposition in the CMMB standard

The China mobile multimedia broadcasting Standard utilizes a highly structured low density parity check code to guarantee the transmission reliability. With the non-systematic and non-quasi-cyclic characteristics of the parity matrix, an encoding algorithm and its compact matrix store scheme are devised. Due to the sparse requirement, an optimized sparse LU decomposition is proposed based on the modified optimization criterion of the minimal product of row and column weights. Computational results show that, comparing to the existing results, our proposed LU decomposition can achieve a 10％ fewer number of the elements “1” in matrices L and U, which contributes the lower complexity of the encoding algorithm. Therefore, the proposed algorithm has a better prospect in the practical CMMB system.