基于贝叶斯框架的OFDM稀疏信道估计算法  

OFDM Sparse Channel Estimation Algorithm Based on Bayesian Framework

作　　者：丁宇舟 颜彪[1] 何豆豆 Ding Yuzhou;Yan Biao;He Doudou(School of Information Engineering,Yangzhou University,Yangzhou Zhejiang 225009,China)

机构地区：[1]扬州大学信息工程学院,江苏扬州225009

出　　处：《山西电子技术》2024年第3期62-65,94,共5页Shanxi Electronic Technology

基　　金：国家自然科学基金项目(61601403)。

摘　　要：为了提高正交频分复用OFDM(Orthogonal Frequency Division Multiplexing,OFDM)稀疏信道估计的性能,实现了一种基于3-L分层先验模型的变分贝叶斯VMP(Variational Message Passing)消息传递算法。该算法对待估计向量的辅助函数分组并且加入贝塞尔函数,通过消息传递原则更新估计值参数,最后估计出频率响应采样值。仿真显示相较于传统的CosaMP、EM算法,提出的变分贝叶斯VMP算法可以获取更好的重构性能。在中高信噪比下,所提出的算法比传统CosaMP、EM算法的误比特率提高了2-3db,均方误差提高了3-4db。In order to improve the performance of orthogonal frequency division multiplexing OFDM(Orthogonal Frequency Division Multiplexing,OFDM)sparse channel estimation,implements a variational Bayes VMP(Variational Message Passing)messaging algorithm based on 3-L hierarchical prior model.The algorithm groups the auxiliary functions of the estimation vector and adds the Bezier function,updates the estimated parameter through the messaging principle,and finally estimates the frequency response sampling value.The simulation shows that compared with the traditional CosaMP and EM algorithms,the proposed variational Bayes VMP algorithm can obtain better reconstruction performance.Under the medium and high signal-to-noise ratio,the proposed algorithm improves the bit error rate by 2-3db and the mean squared error by 3-4db compared with the traditional CosaMP and EM algorithms.

关 键 词：正交频分复用 信道估计 压缩感知 重构算法 变分贝叶斯 

分 类 号：TN911.5[电子电信—通信与信息系统]