基于压缩感知理论的UWB系统信道估计研究

发布时间：2018-05-04 23:00

本文选题：信道估计 + 压缩感知　；参考：《西安电子科技大学》2014年硕士论文

【摘要】：信道估计技术是获取信道状态信息的核心技术之一,依据一定的估计准则实现系统信道对输入信号传输过程产生的影响,进一步达到提高系统性能的目的。而压缩感知理论(CS,Compressed Sensing)最大的优势是以低速率采样(远小于奈奎斯特速率)实现信号重构,而信号重构又是信道估计过程中的关键步骤。应用CS理论需要满足信道和信号稀疏特性,本文在验证超宽带(UWB,Ultra-WideBand)系统信道满足该性质的基础上,利用CS实现该信道的估计过程,从而实现性能优良的信道系统。首先,本文全面介绍压缩感知理论,并阐明其作为信道估计的理论支持,同时分析了超宽带系统的框架及其特点,并验证该系统对于压缩感知理论的适应性,接着在超宽带系统信道模型基础上具体描述了该信道上的信号传输和信号估计过程。其次,给出了一个基于CS的UWB系统信道估计方案。针对传统基于导频的信道估计方案的不足,本文基于压缩感知估计算法提出一个应用于超宽带系统的信道估计方案,仿真结果显示,后者可以实现更好的系统性能且不必付出更多的观测数据,从而也验证了以低于奈奎斯特采样的速率可以实现信号的恢复,仿真还分析了会影响CS恢复算法性能的因素。最后,给出基于SN-CS的UWB系统信道估计方案。针对感知算法中观测矩阵的构造和获取,本文在研究现有高斯随机矩阵和伯努利矩阵作为感知矩阵的基础上提出了一个新的SN-CS(Compressed Sensing Based on Sub-Nyquist)算法。新算法利用基于亚奈奎斯特的宽带调制器架构实现模数转换的低速抽样过程,实现了通过硬件方式直接获取相应的快速傅里叶变换系数来构造感知矩阵,更能贴近信道特性。同时,本文将新算法应用于超宽带系统并提出四种实际的采样结构的框架然后予以仿真实现,并就能量获取和均方根误差两个参数对比本方案和奈奎斯特采样恢复的性能,结果表明本方案中设计结构的实现能够实现几乎和奈奎斯特一样的恢复性能,而且只需要远少于奈奎斯特采样所需要的采样点。
[Abstract]:Channel estimation is one of the key technologies to obtain channel state information, and the influence of the system channel on the input signal transmission process is realized according to certain estimation criteria, and the purpose of improving the system performance is to be further improved. The most important advantage of CS (Compressed Sensing) is to sample at a low rate (far less than Naquis) The signal reconstruction is the key step in the channel estimation. The application of CS theory needs to meet the sparse characteristics of the channel and signal. In this paper, the channel estimation process is realized by using CS to realize the channel estimation on the basis of the authentication of the UWB (Ultra-WideBand) system, so as to realize the channel with excellent performance. Firstly, this paper introduces the theory of compressed sensing, expounds its theoretical support as channel estimation, analyzes the framework and characteristics of UWB system, and verifies the adaptability of the system to the compression perception theory, and then describes the signal transmission and signal on the channel based on the UWB system channel model. Secondly, a channel estimation scheme for UWB system based on CS is given. In view of the shortage of traditional channel estimation based on pilot, this paper proposes a channel estimation scheme applied to UWB based on compressed sensing estimation algorithm. The simulation results show that the latter can achieve better system performance and do not have to pay. More observation data, which also validates the recovery of the signal at a rate below the Nyquist sampling rate. The simulation also analyzes the factors that affect the performance of the CS recovery algorithm. Finally, the SN-CS based channel estimation scheme for the UWB system is given. In view of the construction and acquisition of the observation matrix in the perceptual algorithm, this paper studies the existing Gauss follow. A new SN-CS (Compressed Sensing Based on Sub-Nyquist) algorithm is proposed on the basis of the mechanical matrix and the Bernoulli matrix as the perceptual matrix. The new algorithm uses a broadband modulator architecture based on the sub Nyquist to realize the low speed sampling process of the analog digital conversion, and achieves the corresponding fast Fourier transform directly through the hard piece mode. In this paper, the new algorithm is applied to the UWB system and the framework of four actual sampling structures is proposed and simulated. The performance of the scheme and the Nyquist sampling recovery is compared with the two parameters of energy acquisition and root mean square error. The results show that the scheme is in this scheme. The implementation of the design structure can achieve almost the same recovery performance as Nyquist, and much less than the sampling points needed by Nyquist sampling.

【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN925;TN911.23

【参考文献】