基于导频的MIMO-OFDM无线通信系统信道估计技术研究

发布时间：2018-06-10 01:27

本文选题：多输入多输出正交频分复用 + 导频信道估计　；参考：《南京航空航天大学》2017年硕士论文

【摘要】：随着现代社会对移动通信服务需求的迅速增长,频谱资源日趋紧张,未来可靠高速的无线通信首先要实现大容量、高速率和高频谱利用率。OFDM技术由于频谱利用率高、实现复杂度低、抗多径衰落等优点,在宽带通信中得到广泛的应用。同时,MIMO技术充分利用空间资源,在无需占用额外频谱资源以及增加发射功率的前提下,有效提升系统的信道容量和频谱利用率。因此,MIMO技术与OFDM技术的结合能够同时发挥两种技术的优势,有利于系统容量及频谱利用率的提升。第四代移动通信系统中已经将MIMO-OFDM技术作为数据传输的核心技术方案,很大程度上提升了用户的通信体验。由于MIMO-OFDM无线通信系统中,信道传输环境复杂,因此信道估计是通信系统可靠运行的关键,同时空时解码、相干解调、恢复初始数据等操作也需要尽可能准确的信道状态信息,因此,本文主要研究MIMO-OFDM无线通信系统中的信道估计算法。本文针对MIMO-OFDM系统,首先根据无线信道的统计特性和衰落特性,建立了频率选择性衰落和非频率选择性衰落MIMO信道模型;然后结合OFDM技术利用多载波并行传输信号的特点,建立MIMO-OFDM系统数据传输模型;接着推导了LS、DFT以及MMSE三种经典信道估计算法,均仿真实现。算法改进部分,本文针对DFT算法中最大时延长度内仍存在噪声以及最大时延长度需要提前获取的不足,创新提出了基于循环前缀的改进DFT算法,用循环前缀长度代替最大时延长度,并设置门限值获取循环前缀长度内的信道主要路径,从而提升了经典DFT算法的估计精度;同时,本文针对LS算法和DFT算法结果中每一径上存在噪声的不足,创新提出了基于指数平滑算法的改进LS和DFT信道估计算法,利用指数平滑算法对信道估计的结果做优化,从而减少信道每一径上的噪声,提升估计精度。仿真实验结果表明,基于循环前缀的改进DFT算法提升了经典DFT算法的估计精度,同时避免了经典DFT算法中获取最大时延长度的问题;指数平滑算法对LS、DFT算法的估计结果进行优化,减少信道每一径上的噪声,有效提升了经典LS、DFT算法的估计精度,且改进算法仅需增加少量计算复杂度,易于实现。
[Abstract]:With the rapid growth of the demand for mobile communication services in modern society, the spectrum resources are becoming increasingly tight. In the future, the reliable and high-speed wireless communication must first achieve large capacity, high rate and high spectrum efficiency. OFDM technology is due to the high spectrum efficiency. Because of its low complexity and anti-multipath fading, it has been widely used in broadband communication. At the same time MIMO technology makes full use of space resources and improves the channel capacity and spectral efficiency of the system without requiring additional spectrum resources and increasing transmission power. Therefore, the combination of MIMO technology and OFDM technology can bring into play the advantages of the two technologies at the same time, which is conducive to the improvement of system capacity and spectrum efficiency. MIMO-OFDM technology has been taken as the core technology of data transmission in the fourth generation mobile communication system, which greatly improves the user's communication experience. In MIMO-OFDM wireless communication system, the channel transmission environment is complex, so channel estimation is the key to reliable operation of communication system. At the same time, space-time decoding, coherent demodulation, recovery of initial data and other operations also need as accurate channel state information as possible. Therefore, this paper mainly studies the channel estimation algorithm in MIMO-OFDM wireless communication systems. For MIMO-OFDM system, firstly, according to the statistical and fading characteristics of wireless channel, the MIMO channel models of frequency-selective fading and non-frequency-selective fading are established, and then the characteristics of multi-carrier parallel transmission signal are combined with OFDM technology. The data transmission model of MIMO-OFDM system is established, and then three classical channel estimation algorithms, LS- DFT and MMSE, are derived and realized by simulation. In order to solve the problem that the maximum delay length still exists in the DFT algorithm and the maximum delay length needs to be obtained in advance, an improved DFT algorithm based on cyclic prefix is proposed in this paper. The maximum delay length is replaced by cyclic prefix length, and the main channel path within the cyclic prefix length is obtained by setting the threshold value, which improves the estimation accuracy of the classical DFT algorithm. Aiming at the deficiency of noise in every path of LS algorithm and DFT algorithm, an improved LS and DFT channel estimation algorithm based on exponential smoothing algorithm is proposed in this paper, and the result of channel estimation is optimized by exponential smoothing algorithm. In order to reduce the channel noise on each path, improve the estimation accuracy. The simulation results show that the improved DFT algorithm based on cyclic prefix improves the estimation accuracy of the classical DFT algorithm and avoids the problem of obtaining the maximum delay length in the classical DFT algorithm. The exponential smoothing algorithm optimizes the estimation results of the LSN DFT algorithm, reduces the noise in every path of the channel, effectively improves the estimation accuracy of the classical LSN DFT algorithm, and the improved algorithm only needs a small amount of computational complexity, which is easy to implement.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.5

【参考文献】