摘要
针对采用低密度奇偶校验(LDPC)编码的多输入多输出系统中检测与译码级联方案未联合优化的问题,提出了一种低复杂度联合贝尔实验室分层空时(VBLAST)最小均方误差迭代检测与LDPC译码算法.该算法将VBLAST检测和LDPC译码统一于一个迭代信号处理框架,从系统角度分析了整个联合译码检测;通过级联简化的硬判决译码与检测的混合迭代消息传播,减少了各模块单独处理引起的互信息处理损失,从而优化了检测性能并降低了复杂度.仿真结果表明:与全软信息辅助检测方案相比,它以微小的0.1~0.2dB性能损失代价,获得了前者约1/2的复杂度及检测延迟等优势.
According to the non-joint-optimization problem of independent detection and decoding separately in a low-density parity-check(LDPC)coded multiple-input multiple-output systems,a low complex algorithm of jointly iterative vertical Bell laboratories layered space-time(VBLAST)minimum mean squared error detection and LDPC decoding was proposed.The whole system of the joint decoding and detection was analyzed in the systematic aspect,where it was united in the same iterative signal processing framework.Also,the detection performance of the system was optimized and the complexity was reduced by the hybrid iterative message propagation with the concatenation of simplified hard decoding and detection for optimized performance,which reduced the mutual information processing loss among all separated functional parts.Simulation results indicate that the proposed algorithm can obtain better advantages of about half lower complexity of decoding and detection and half less iterative processing latency at the cost of slightly 0.1~0.2dB performance loss,when compared with all soft metric based on detection scheme.
出处
《华中科技大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2015年第10期58-63,共6页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国家自然科学基金资助项目(61320106003
61471152)
浙江省自然科学基金资助项目(LZ14F010003)
国家高技术研究发展计划资助项目(2012AA01A506)
东南大学移动通信国家重点实验室开放研究基金资助项目(2014D02)
中国博士后科学基金资助项目(2014M561555)
关键词
多输入多输出系统
迭代译码检测
LDPC译码
VBLAST检测
最小均方误差
外信息转移图
multiple-input multiple-output systems
iterative decoding and detection
low-density parity-check(LDPC)decoding
vertical Bell laboratories layered space-time(VBLAST) detection
minimum mean squared error(MMSE)
EXIT chart