OFDM稀疏信道估计中的导频优化研究

发布时间：2018-07-07 20:19

本文选题：压缩感知 + OFDM　；参考：《电子科技大学》2017年硕士论文

【摘要】：近年来,随着压缩感知(Compressed sensing,CS)理论的普及,该技术已经被广泛应用于正交频分复用(Orthogonal Frequency Division Multiplexing,OFDM)的稀疏信道估计研究中。不同于传统的信道估计方法,基于CS的信道估计技术可以利用极低的采样速率对信号进行采样,并有效地重建信号,这样可以减少对导频的使用,提高系统的传输效率。其中,常见的稀疏信号重构算法主要有:正交匹配追踪(Orthogonal matching pursuit,OMP)算法,压缩采样匹配追踪(Compressive sampling matching pursuit,CoSaMP)算法等贪婪算法。此外,一些凸优化算法如SpaRSA(Sparse Reconstruction by Separable Approximation)和YALL1等算法同样也可以作为重构算法对稀疏信号进行恢复。然而,大部分学者把研究的重点都放在了信道估计算法的改进与创新上,却忽略了影响信道估计性能的其他因素,如导频结构的设计。现有研究表明,不同的导频结构对稀疏信道估计的最终性能也起到了十分重要的作用。因此,本文将对OFDM稀疏信道估计下的导频结构设计问题进行重点研究,通过对导频结构进行针对性设计,提高整个系统的信道估计性能。本文首先在确定性导频结构设计标准的基础上,对现有标准及其实现算法进行了归纳与总结,并在此基础上实现了完善与改进,减小了算法的复杂度,提高了算法的收敛速率。此外,本文还给出了一种自适应导频结构优化设计算法。不同于传统的确定性导频设计算法,该自适应算法将根据实际传输环境对导频结构进行动态的调整,使整个系统始终保持较高的估计性能。本文的主要工作如下:1.本文对现有的确定性导频结构设计标准进行了归纳与总结,并针对不同的标准分别给出了一种具体的算法实现。通过对仿真实验结果的分析与对比,详细说明了各标准所适用的场景以及估计性能之间的差异。2.本文在传统MIP标准的基础上对其实现算法进行了改进,通过与遗传算法的结合,降低了原有算法的复杂度,提高了算法的稳定性及收敛速率。3.本文对MIP标准本身进行了完善,给出了一种改进后的确定性导频结构设计标准。与一般的MIP标准相比,改进后的标准更充分地考虑了稀疏信号在恢复过程中的其他因素,丰富了对测量矩阵的设计,使在该标准下所得到的导频结构具有更加稳定更加精确的估计性能。4.本文提出了一种自适应导频结构设计思想,并给出了其具体的算法实现。不同于传统的确定性导频结构设计,该思想强调将导频结构的设计问题与实时的信道估计相结合,利用实时估计出的稀疏信道脉冲响应(Channel impulse response,CIR)反馈到导频结构的重构设计上。此外,为了解决贪婪算法在恢复精度上欠缺的问题,本文还利用了凸优化算法中的最小1l范数模型对导频结构进行进一步的筛选,并模拟遗传算法中的迭代过程不断地对导频结构进行重建,直至其最终收敛。5.为了降低自适应算法在工程应用中的实现难度,本文还将确定性导频结构设计算法与自适应性导频结构设计算法相结合,通过对导频结构进行一定的预处理,减少了自适应性算法所需要的收敛时间,提高了算法的效率。
[Abstract]:In recent years, with the popularity of Compressed sensing (CS) theory, this technology has been widely used in the research of sparse channel estimation for orthogonal frequency division multiplexing (Orthogonal Frequency Division Multiplexing, OFDM). Unlike traditional channel estimation methods, the channel estimation based on CS can make use of very low sampling rates. The signal is sampled and the signal is rebuilt effectively, which can reduce the use of pilot and improve the transmission efficiency of the system. Among them, the common sparse signal reconstruction algorithms are orthogonal matching tracking (Orthogonal matching pursuit, OMP), compressed sampling matching tracking (Compressive sampling matching pursuit, CoSaMP) and so on In addition, some convex optimization algorithms such as SpaRSA (Sparse Reconstruction by Separable Approximation) and YALL1 can also be used as reconstruction algorithms to restore sparse signals. However, most of the scholars focus on the improvement and innovation of the channel estimation method, but ignore the influence of channel estimation. Other factors, such as the design of pilot structures, have shown that different pilot structures play a very important role in the final performance of sparse channel estimation. Therefore, this paper will focus on the design of pilot structure under OFDM sparse channel estimation, and improve the pilot structure to improve the pilot structure. The performance of the channel estimation for the whole system. Firstly, based on the design standard of deterministic pilot structure, the existing standard and its implementation algorithm are summarized and summarized. On this basis, the improvement and improvement are realized, the complexity of the algorithm is reduced and the convergence rate of the algorithm is improved. In addition, this paper also gives an adaptive guide. The frequency structure optimization design algorithm is different from the traditional deterministic pilot design algorithm. The adaptive algorithm will dynamically adjust the pilot structure according to the actual transmission environment, and make the whole system maintain a high estimation performance. The main work of this paper is as follows: 1. this paper has carried out the standard of certain deterministic pilot structure design. A specific algorithm implementation is given for different standards. Through the analysis and comparison of the results of the simulation experiments, the scene and the difference between the estimated performance are explained in detail..2. is improved on the basis of the traditional MIP standard and through the genetic algorithm. Combining it, the complexity of the original algorithm is reduced, the stability and the convergence rate of the algorithm are improved.3.. The MIP standard itself is perfected in this paper. A modified deterministic pilot structure design standard is given. Compared with the general MIP standard, the improved standard more fully considers the other factors of the sparse signal in the recovery process. The design of the measurement matrix is enriched and the pilot structure obtained under this standard has a more stable and more accurate estimation performance..4. this paper proposes an adaptive pilot structure design idea and gives its specific algorithm implementation. Combined with real-time channel estimation, the real-time estimated sparse channel impulse response (Channel impulse response, CIR) is used to feed back to the pilot structure reconfiguration design. In addition, in order to solve the problem that the greedy algorithm lacks the recovery precision, this paper also uses the minimum 1L norm model in the convex optimization algorithm for the pilot junction. Further screening is carried out, and the iterative process in the genetic algorithm is simulated continuously to reconstruct the pilot structure until its final convergence is.5. in order to reduce the difficulty of realizing the adaptive algorithm in the engineering application. The preprocessing of the structure reduces the convergence time of the adaptive algorithm and improves the efficiency of the algorithm.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.53

【参考文献】