摘要
基于无线通信中接收信号的似然检测模型,给出发送信号估计和推理的因素图描述。在此基础上,应用标准和积算法提出一种频率选择性衰落信道下的多输入多输出(multiple input multiple output,MIMO)系统的迭代信号检测算法。为了降低计算复杂度,提出一种快速迭代检测算法,通过引入概率数据关联方法,避免了标准和积算法中符号概率估计的积分计算。仿真结果表明,应用标准和积算法得到的迭代信号检测算法可以逼近最优算法的性能,而快速迭代检测算法在大幅降低计算复杂度的同时,误码率性能比传统的最小均方误差算法有较大优势。
A factor graph description of transmitted signal estimation and inference is presented based on the likelihood detection model of the received signal in wireless communications. The sum-product algorithm is applied to the factor graph for iterative signal detection over the frequency selective channel in MIMO systems. Furthermore, a fast iterative detection algorithm is proposed to reduce the computational complexity, where a probabilistic data association method is introduced to avoid integral computation in the normal sum-product algorithm. Simulation results show that the proposed iterative signal detection algorithm using the normal sumproduct algorithm approximates to optimum performance. The computational complexity of the fast iterative algorithm can be greatly reduced; meanwhile its bit error rate performance is also better than the traditional mini- mum mean square error algorithm.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2009年第7期1534-1538,共5页
Systems Engineering and Electronics
基金
国家自然科学基金资助课题(60672047)
关键词
多输入多输出
信号检测
因素图
和积算法
概率数据关联
multiple input multiple output
signal detection
factor graph
sum-product algorithm
probabilistic data association