基于压缩感知的MIMO-OFDM系统信道估计算法研究

发布时间：2018-06-20 04:20

本文选题：MIMO-OFDM + 信道估计　；参考：《兰州交通大学》2016年硕士论文

【摘要】：自人类诞生开始,我们从未放弃对通信技术的研究。在2004年12月的3GPP多伦多会议上,LTE(Long Term Evolution,长期演进)正式立项并启动,标志着第四代移动通信(4G)技术的时代已经到来。LTE是3G的演进技术,采用OFDM(正交频分复用)技术和MIMO(多输入多输出)技术作为唯一标准,堪称目前民用移动通信领域的顶尖技术。MIMO-OFDM无线通信系统凭借高信道容量和高频谱利用率已经被广泛地应用于需要高速、大数据量的通信场合。MIMO-OFDM系统优势显著,但缺点也不容忽视,那就是在信息的获取环节极其依赖信道的状态信息参数。所以信道估计作为MIMO-OFDM无线通信系统中的关键步骤,算法的好坏会直接影响到整个系统的性能表现。所谓信道估计,就是利用导频对无线信道进行采样,并在系统接收端应用相应的恢复算法计算出信道参数。信道估计效果的好坏,决定着信息传递质量的高低。无线信道环境非常复杂,参数并不像有线信道那样固定可预见。在某些环境下,缺少信道估计环节的无线通信系统的性能会受到很大影响。在传统的MIMO-OFDM系统信道估计算法中,导频的放置均必须满足奈奎斯特定理。对于快速变化的无线信道,这意味着系统需要传输大量的导频信号,占用大量的频带资源。尽管OFDM技术可以提高系统的频谱效率,但如果节省下的频带资源被不携带任何信息的导频信号占用,节省出来的频段将被白白浪费,这与MIMO-OFDM系统设计的初衷背道而驰。压缩感知理论指出,若被采样信号是稀疏的,那么就能够以远低于奈奎斯特定理所需的采样值个数采样信号,并通过重构算法精确地恢复出原始信号。本文研究的最终目的是减少信道估计所需传输的导频数量,节省系统的频带资源,同时不影响系统的整体性能。由于大多数MIMO-OFDM信道具有稀疏或近似稀疏的信道参数,因此本文应用压缩感知理论代替奈奎斯特定理对信道参数进行采样,并运用压缩感知重构算法重构信道参数。为了在确保估计性能的同时弥补随机导频的不足,本文设计了一种确定导频来代替传统的均匀分布和随机导频。仿真结果表明,确定导频与随机导频保持了相同的信道估计性能。虽然在传统的MIMO-OFDM信道估计中,均匀分布的导频可以为系统带来最佳的性能表现,但在基于压缩感知的算法中,均匀分布的导频却不能得到很好的信道估计效果。最后的仿真实验结果显示,基于压缩感知的MIMO-OFDM信道估计算法不仅可以使得所需传送的导频数量大大减少,同时又可以确保高精确度的估计性能,达到本论文题目的研究目的。
[Abstract]:Since the birth of mankind, we have never given up our research on communication technology. At the 3GPP Toronto Conference in December 2004, LTEL long term Evolution (long term Evolution) was formally proposed and launched, indicating that the era of the fourth generation mobile communication 4G) technology has arrived. LTE is the evolution technology of 3G. OFDM (orthogonal Frequency Division Multiplexing) and Mimo (multiple input and multiple output) technologies are used as the sole standard. MIMO-OFDM wireless communication system, which can be regarded as the top technology in the field of civil mobile communication at present, has been widely used in communication field with high channel capacity and high spectral efficiency, which requires high speed and large amount of data. MIMO-OFDM system has a significant advantage. However, the disadvantage can not be ignored, that is, the acquisition of information is extremely dependent on the channel state information parameters. Therefore, as a key step in MIMO-OFDM wireless communication system, channel estimation will directly affect the performance of the whole system. The so-called channel estimation is to sample the wireless channel by using pilot frequency and calculate the channel parameters by using the corresponding recovery algorithm at the receiver of the system. The quality of channel estimation determines the quality of information transmission. The wireless channel environment is very complex, the parameters are not as fixed and predictable as the wired channel. In some environments, the performance of wireless communication systems without channel estimation will be greatly affected. In the traditional channel estimation algorithms for MIMO-OFDM systems, pilot placement must satisfy Nyquist's theorem. For rapidly changing wireless channels, this means that the system needs to transmit a large number of pilot signals and occupy a large amount of frequency band resources. Although OFDM technology can improve the spectral efficiency of the system, if the saved frequency band resources are occupied by pilot signal without any information, the saved frequency band will be wasted, which is contrary to the original intention of MIMO-OFDM system design. Compression sensing theory points out that if the sampled signal is sparse, the sample signal can be sampled at a number far below the sampling value required by Nyquist's theorem, and the original signal can be accurately recovered by the reconstruction algorithm. The ultimate purpose of this paper is to reduce the number of pilots needed for channel estimation and to save the frequency band resources of the system without affecting the overall performance of the system. Because most MIMO-OFDM channels have sparse or nearly sparse channel parameters, compression sensing theory is used instead of Nyquist theorem to sample channel parameters, and compression perception reconstruction algorithm is used to reconstruct channel parameters. In order to ensure the performance of estimation and make up for the deficiency of random pilot, a new method of determining pilot is designed to replace the traditional uniform distribution and random pilot. The simulation results show that the performance of channel estimation is the same as that of random pilot. Although in the traditional MIMO-OFDM channel estimation, the uniformly distributed pilot can bring the best performance for the system, but in the compressed sensing algorithm, the uniformly distributed pilot can not get a good channel estimation effect. Finally, the simulation results show that the compressed sensing based MIMO-OFDM channel estimation algorithm can not only greatly reduce the number of pilots to be transmitted, but also ensure the high accuracy of the estimation performance.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN919.3

【参考文献】