OFDM系统中基于压缩感知的信道估计研究
发布时间：2018-10-29 10:22
【摘要】：本文研究了基于OFDM系统的压缩感知信道估计中导频模式的选取和DDLMMSE信道估计算法。建立了压缩感知应用于信道估计中的数学模型,并对压缩感知信道估计中如何选取最优的导频模式进行了研究,提出了一种改进的选取方法,通过仿真验证了结果;阐明了DDLMMSE信道估计算法存在的不足并提出了改进算法,仿真实验验证了改进算法的优越性。主要研究内容包括:(1)提出了一种改进的压缩感知信道估计导频选取方法。理论分析和仿真实验证明了利用相关性衡量观测矩阵并用其最小准则选取导频模式存在的缺陷,因而提出利用加权相关函数来代替相关性,弥补了利用相关性选取导频模式的缺陷。将用加权相关函数选出的导频模式进行信道估计,利用OMP及LS算法重构出信道响应。100次Monte Carlo实验结果表明,利用加权相关函数选则的导频模式进行信道估计的结果中,30%优于相关性选择的导频模式,68%与相关性选取的导频模式相同,2%的其他情况;通过对加权相关函数的仿真得出,当总的循环次数在300次左右时,加权相关趋于收敛,因此将循环次数设为300,减少仿真运行的时间。(2)提出了一种改进的DDLMMSE信道估计算法。DDLMMSE信道估计算法是对LS算法进行修正所得,改进的DDLMMSE信道估计算法利用子载波间的相关性,将DDLMMSE算法中的对角阵扩展到K重矩阵,在不显著增加算法复杂度的前提下,提高信道估计的精度。仿真表明,当误码率一定时,随着K的增加需要的信噪比减小,K为3时比K为1时所需的信噪比约减小0.5d B,因此,可以根据对性能的要求来选取合适的K。
[Abstract]:In this paper, pilot mode selection and DDLMMSE channel estimation algorithm in compressed perceptual channel estimation based on OFDM system are studied. The mathematical model of compressed sensing applied to channel estimation is established, and how to select the optimal pilot mode in compressed perceptual channel estimation is studied. An improved selection method is proposed, and the results are verified by simulation. The shortcomings of the DDLMMSE channel estimation algorithm are clarified and the improved algorithm is proposed. The simulation results demonstrate the superiority of the improved algorithm. The main contents are as follows: (1) an improved pilot selection method for compressed perceptual channel estimation is proposed. Theoretical analysis and simulation experiments prove the defects of selecting pilot mode by using correlation measurement observation matrix and its minimum criterion. Therefore, a weighted correlation function is proposed to replace correlation. It makes up for the defect of selecting pilot mode by correlation. The pilot mode selected by weighted correlation function is used for channel estimation, and the channel response is reconstructed by using OMP and LS algorithms. The results of 100 Monte Carlo experiments show that the pilot mode of weighted correlation function selection is used for channel estimation. 30% is superior to the pilot mode selected by correlation, 68% is the same as the pilot mode selected by correlation, and 2% is other cases. Through the simulation of the weighted correlation function, when the total number of cycles is about 300 times, the weighted correlation tends to converge, so the number of cycles is set to 300, (2) an improved DDLMMSE channel estimation algorithm is proposed. The DDLMMSE channel estimation algorithm is a modified LS algorithm, and the improved DDLMMSE channel estimation algorithm utilizes the correlation between subcarriers. The diagonal matrix in the DDLMMSE algorithm is extended to the K-multiplicity matrix, which improves the accuracy of channel estimation without significantly increasing the complexity of the algorithm. Simulation results show that when the bit error rate is constant, the SNR required by K decreases with the increase of K, and the SNR required by K at 3 is about 0.5 dB lower than that required by K at 1. Therefore, appropriate Ks can be selected according to the performance requirements.
【学位授予单位】：天津理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN929.53
本文编号：2297420
[Abstract]:In this paper, pilot mode selection and DDLMMSE channel estimation algorithm in compressed perceptual channel estimation based on OFDM system are studied. The mathematical model of compressed sensing applied to channel estimation is established, and how to select the optimal pilot mode in compressed perceptual channel estimation is studied. An improved selection method is proposed, and the results are verified by simulation. The shortcomings of the DDLMMSE channel estimation algorithm are clarified and the improved algorithm is proposed. The simulation results demonstrate the superiority of the improved algorithm. The main contents are as follows: (1) an improved pilot selection method for compressed perceptual channel estimation is proposed. Theoretical analysis and simulation experiments prove the defects of selecting pilot mode by using correlation measurement observation matrix and its minimum criterion. Therefore, a weighted correlation function is proposed to replace correlation. It makes up for the defect of selecting pilot mode by correlation. The pilot mode selected by weighted correlation function is used for channel estimation, and the channel response is reconstructed by using OMP and LS algorithms. The results of 100 Monte Carlo experiments show that the pilot mode of weighted correlation function selection is used for channel estimation. 30% is superior to the pilot mode selected by correlation, 68% is the same as the pilot mode selected by correlation, and 2% is other cases. Through the simulation of the weighted correlation function, when the total number of cycles is about 300 times, the weighted correlation tends to converge, so the number of cycles is set to 300, (2) an improved DDLMMSE channel estimation algorithm is proposed. The DDLMMSE channel estimation algorithm is a modified LS algorithm, and the improved DDLMMSE channel estimation algorithm utilizes the correlation between subcarriers. The diagonal matrix in the DDLMMSE algorithm is extended to the K-multiplicity matrix, which improves the accuracy of channel estimation without significantly increasing the complexity of the algorithm. Simulation results show that when the bit error rate is constant, the SNR required by K decreases with the increase of K, and the SNR required by K at 3 is about 0.5 dB lower than that required by K at 1. Therefore, appropriate Ks can be selected according to the performance requirements.
【学位授予单位】：天津理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN929.53
【参考文献】
相关期刊论文 前1条
1 张建康;陈恩庆;穆晓敏;齐林;;OFDM系统中一种最大似然信道估计算法[J];兵工学报;2009年09期
，本文编号：2297420
本文链接：https://www.wllwen.com/kejilunwen/wltx/2297420.html
