| 摘要: |
| 在高速通信系统中，由于多径信道通常存在一些小的散射体，使得抽头向量不满足理想的稀疏特性，导致经典的稀疏估计算法存在一定的性能损失。针对上述非理想稀疏特性问题，提出了一种基于酉变换近似消息传递（Unitary Transform Approximate Message Passing，UT-AMP）和加权高斯（Weighting-Gaussian,WG）先验模型的稀疏估计算法。首先，由非理想稀疏信道的构造分析，导出了WG先验模型和参数；其次，利用贝叶斯公式对正交频分复用(Orthogonal Frequency Division Multiplexing,OFDM)系统进行因式分解和因子图建模，归纳了在消息传递框架内期望最大化（Expectation Maximization，EM）算法嵌入方式，推导了联合UT-AMP和EM的信道估计算法；最后，建立仿真环境对所提算法进行复杂度分析和数值仿真。仿真结果表明，所提算法能够以同阶复杂度实现信道估计性能和频带利用率的提升，具有很高的应用和推广价值。 |
| 关键词: 非理想稀疏信道 信道估计 期望最大化算法 加权高斯先验 压缩感知 |
| DOI： |
|
| 基金项目:国家自然科学基金资助项目(61801434)；中国博士后科学基金面上项目(2019M652576)；河南省科技公关项目(202102210313,202102210172,202102210556)；河南省高等学校重点研究课题(20B510005) |
|
| Non-ideal sparse channel estimation using weighting-Gaussian model |
| GAO Tongdi,QIN Xuezhen,YUAN Zhengdao,WANG Jiabin |
| (1.Artificial Intelligence Engineering Research Center,Open University of Henan,Zhengzhou 450001,China;2.School of Information Engineering,Zhengzhou University,Zhengzhou 450001,China;3.The 713th Research Institute of China Shipbuilding Industry Corporation,Zhengzhou 450001,China) |
| Abstract: |
| In high speed communication systems,there are some small scatterers in multipath channel,so that the tap vector does not have ideal sparse characteristics,leading to some performance loss to sparse estimation algorithm.A sparse estimation algorithm based on unitary transform approximate message passing (UT-AMP) and weighting-Gaussian(WG) prior model is proposed to solve the such problem.The algorithm firstly analyzes the construction of non-ideal sparse channel and introduces the WG prior model.Secondly,the orthogonal frequency division multiplexing(OFDM) system is factorized and modeled by using Bayesian theorem,the summarization of the embedding of expectation maximization（EM） into message passing framework is studied,and then the channel estimation algorithm of joint UT-AMP and EM is derived.Finally,a simulation environment is built for complexity analysis and numerical simulation.Simulation results show that,the proposed algorithm can improve the performance of channel estimation and spectrum efficiency with the same order complexity,which is of high application value. |
| Key words: non-ideal sparse channel channel estimation expectation maximization weighting-Gaussian prior compressed sensing |
