MIMO-OFDM信道估计算法研究

【作者】 王钢飞；

【导师】 卓东风；

【作者基本信息】 太原科技大学 ， 电路与系统， 2011， 硕士

【摘要】 正交频分复用技术具有高速传输数据、频谱利用率高、抗频率选择性衰落和多径衰落的特点,而多输入多输出技术具有在不增加系统带宽的情况下提高通信系统的容量的特点。MIMO-OFDM系统具备两种技术的优点,能同时获得时间、频率和空间三种分集,是下一代(4G)无线通信的关键技术。信道估计技术是MIMO-OFDM系统发挥其优越性的关键技术,因此有必要对MIMO-OFDM系统的信道估计方法进行研究。首先,本文介绍了OFDM系统、MIMO系统和MIMO-OFDM系统的基本原理和结构,以及无线信道的空间特点和衰落特性,研究了基于导频符号的信道估计算法,主要是最小二乘法和奇异值分解法。同时,研究了线性插值、高斯插值和拉格朗日插值等常用插值算法,通过计算机仿真得到了性能对比曲线。其次,针对频率选择性衰落信道下MIMO-OFDM的系统,基于瑞利衰落信道的模型,利用m序列的自相关性,提出了一种基于m序列做梳状导频的多输入多输出正交频分复用最小二乘算法,该算法可以避免对大矩阵求伪逆,以减少计算复杂度,从而提高了信道估计的计算性能。最后,介绍了叠加导频的MIMO-OFDM系统,采用正交多相序列作为叠加导频,利用导频符号的自相关及导频和接收信号的互相关,并且提出了一种选择性叠加导频信道估计信道的算法。最后,通过理论分析和计算机仿真,表明该方法在快衰落环境中有良好的性能。更多还原

【Abstract】 OFDM has the characteristic of high-speed data transmission, high spectrum efficiency, selective fading with resisted frequency and multipath fading, while multiple input multiple output technology can improve communication systems capacity in case of not increasing system bandwidth. MIMO-OFDM is the key technology of next generation (4G) wireless communication with the advantages of both technologies, it can obtain three diversities of time, frequency and spatial at the same time. Channel estimation plays key roles in MIMO-OFDM system, it is essential to study the channel estimation method of MIMO-OFDM system.First, the article describes the basic principle and structure of OFDM systems, MIMO systems, MIMO-OFDM system, also, the spatial characteristic and fading feature of wireless channel, also, studies the channel estimation algorithm based on pilot symbols as for the least squares method and the singular value decomposition method concretely. Also, the article studies the commonly used interpolation algorithm such as linear interpolation, Gauss interpolation and Lagrange interpolation, getting the comparison curve of performance through computer simulation.Secondly, Focusing on MIMO-OFDM system under frequency selective fading channel, based on the model of Rayleigh fading channel and the autocorrelation of m sequence, the paper proposes a MIMO-OFDM least squares algorithm using m sequence as comb pilot, which avoids pseudo inverse of large matrix so as to reduce the computational complexity and improves the calculation performance of channel estimation.Finally, the paper introduces the MIMO-OFDM system with superimposed pilot, adopting the orthogonal multi-phase sequence as superimposed pilot, using the autocorrelation of pilot symbols, also the cross correlation of pilot and received signal, we propose a channel estimation algorithm based on selective superimposed pilot channel. Theoretical analysis and computer simulation show that the method has good performance in fast fading environment.更多还原