LDPC Block-Coding Schedule Based on Queen-Matrix
Xiling Luo, Jiayi Zhang, Jun Zhang · 2008
The general coding schedule of low density parity codes (LDPC) have the problems of slow coding speed and difficult implementation in hardware. Based on the irregular Queen-matrix LDPC construction algorithm derived from n- queens problem, this paper designed an optimized block-coding schedule according to construction codes' dimension and memory's bandwidth. The block-coding schedule can reduce the intermedial size of data storages and be propitious to hardware implementation by using data source and generating matrix' s division. The experimental results show that comparing with the general coding schedule, the block-coding schedule can make coding speed be improved by 14.6%~36.1% at the cost of a little increasing of bit error ratio (BER) because it arranges the sub-matrix multiplication in parallel prior to sum of them.