摘要
通过分析节点的信息分布和量化饱和效应,提出一种基于伸缩因子的量化方案.该方案不但对校验节点和变量节点的信息进行伸缩,同时还对初始化阶段的信道信息进行伸缩处理.通过对节点的信息收缩,该方案可以使信息的整数部分表示更多的有效信息,同时也降低量化饱和的影响.该文采用(q,0)对节点进行量化,由于节点信息更新全部采用整数而不需要浮点操作,大大减少了硬件复杂性.同时,对(7493,3048)的LDPC码字仿真表明,该量化方案的性能与浮点性能仅相差0.1 dB.
This paper proposes a quantization scheme for low density parity check(LDPC) codes based on scaling factor by analyzing message distribution and the effect of quantization saturation. Messages are scaled for the check nodes and variable nodes, and for the initial channel value. Using the scaling, the scheme allows the integer part of message to include more information, and can lower the effect of quantization saturation. Messages of nodes are quantized with (q, 0). Due to the application of integer part instead of floating point operation, the hardware complexity is significantly reduced. Simulation of (7493, 3048) using normalized Min-Sum displays an implementation loss of about 0.1 dB compared with floating performance.
出处
《应用科学学报》
EI
CAS
CSCD
北大核心
2011年第1期33-38,共6页
Journal of Applied Sciences
基金
上海科委AM基金(No.08700741200)
上海大学新型显示技术与应用集成教育部重点实验室开放课题基金(No.P200802)资助
关键词
量化
伸缩因子
信息分布
量化饱和
LDPC
quantization, scaling factor, message distribution, quantization saturation, LDPC