文档介绍：代号 10701 学号 1008120441
分类号 TN911 密级公开
题(中、英文) 目压缩频谱感知算法研究
Study pressive Spectrum Sensing
作者姓名文璐指导教师姓名、职务赵林靖副教授
学科门类工学学科、专业通信与信息系统
提交论文日期二零一三年三月
摘要
压缩频谱感知主要是指基于压缩感知理论的频谱感知技术。本文主要研究了
压缩感知的基本理论以及压缩频谱感知的相关算法。首先简单介绍了认知无线电
理论以及相关研究的发展现状。之后介绍了频谱感知问题的数学模型,并且对现
有的较为成熟的频谱感知算法进行描述和总结,分析了各种方法的优缺点。然后
对压缩感知理论进行描述,在研究了基于最大最小特征值的感知算法之后,将压
缩感知理论与基于最大最小特征值的方法结合,提出一种压缩频谱感知方法,并
且在信噪比不确定的条件下根据算法的检测概率、虚警概率以及重构误差方面评
估了算法性能,并与同条件下的能量检测算法进行比较,体现出性能上的优势。
之后介绍了一种基于自相关矩阵重构的功率谱重构方法,并在这种方法的支撑下
改进了前文所述的压缩频谱感知算法,通过仿真可以看出,改进算法在重构误差
和检测性能上均优于基于最大最小特征值的压缩频谱感知方法,同时重构的算法
复杂度更低。接着,将这种算法推广到多个物理信道的感知场景下,并通过仿真
验证了算法可用性。最后,利用多台 USRP 搭建实验环境,模拟了具有多个物理
信道的场景,利用采集获得的实测数据对上述算法进行仿真和测试,并在 USRP
上实现提出的算法。
关键词:频谱感知压缩感知自相关矩阵重构最大-最小特征值检测
Abstract
Compressive spectrum sensing refers to spectrum sensing technology based on
compressive sensing. In this paper we mainly focus pressive sensing and
compressive spectrum sensing algorithms. First of all, a brief introduction to concept of
cognitive radio and its recent development is made. Then, the mathematic model of
spectrum sensing and some mature sensing algorithms are described, and their
advantages and disadvantages are summarized. As followed,the basic theory of
compressive sensing is studied, furthermore, bined it with the sensing method
based on max-min eigenvalues as an pressive sensing algorithm. A
simulation is made to evaluate the performance of the proposed method in terms of the
distortion, the probability of detection and false alarm, under the assumption that the
SNR is uncertain. Compared with energy detection, the simulation result shows the
proposed algorithm performs better. And then, a reconstruction of autocorrelation
matrix algorithm is utilized in a modified algorithm to improve the method proposed
above. According to the following simulation, it shows that the modified algorithm can
achieve much lower distortion