摘要
压缩感知(compressed sensing)理论是近两年信号处理领域方兴未艾的一个热门研究方向,它的出现突破了奈奎斯特采样定理对信号频率的限制。最小l1范数法(即BP算法)是压缩感知中信号若干种重构方法中比较成熟的一种算法。文中在对已有的BP算法进行了研究后,提出了一种适用于噪声环境下的加权迭代l1算法,并将之应用于认知无线电的频谱感知中,经过MATLAB仿真的对比后,验证了改进之后的算法对提高含有高斯白噪声的信号重构的精确度有着更加精确的效果。
In recent years, compressed sensing theory is a hot research direction in signal processing field. It breaks through the limit of the Nyquist sampling frequency of signal. Minimum l1 norm method ( ie BP algorithm) is a mature reconstruction method in compressed sensing. This paper proposes a method which changes the noise limit condition from 12 norm to l1 norm and uses iterative weighted l1 norm. The method is applied in the spectrum sensing of cognitive radio. MATLAB simulation results show that the new meth- od reconstructed signal accurately in the AWGN channel.
出处
《南京邮电大学学报(自然科学版)》
北大核心
2012年第3期5-9,15,共6页
Journal of Nanjing University of Posts and Telecommunications:Natural Science Edition
基金
国家重点基础研究发展计划(973计划)(2011CB302903)
国家自然科学基金(60971129)资助项目
关键词
压缩感知
最小l1范数
信号重构
认知无线电
频谱感知
compressed sensing
minimum lI norm
signal reconstruction
cognitive radio
spectrum sensing