摘要
传统的频谱感知方式因其自身的局限性而难以满足处理速率的更高要求。压缩感知的优势在于前端传感器采样数据量远远小于传统采样方法所获的数据量,从而有效提高频谱感知的速度。本文给出了一种新颖的压缩感知算法——快速傅里叶采样算法,该算法能采集较少的点数,较快地重构信号。算法分为频率分离、比特测试和系数估计三个主要步骤。文中对FFS算法进行了详细分析及实现,给出了仿真信`号重构结果 ,并将该算法的运算量与快速傅里叶变换进行了对比分析。仿真结果表明,该算法具有较好的重构精度,并且信号长度的剧烈增加对运算量的影响较小,在大信号处理时运行时间远远低于FFT。
Due to its own limitation, it is difficult to meet higher demands about processing speed. Compress sensing has advantage that its sample data amount is far smaller than data amount acquired by traditional sample way. So speed of spectrum sensing is increased effectively. This paper gives a novel compress sensing algorithm-Fast Flourier Sample（ FFS） algorithm. It can reconstruct signal faster by sample smaller data. Algorithm is mainly divided into three steps, frequency shattering, bit testing and coefficient estimation. This paper analyzes algorithm theory in detail, gives a reconstruct result about simulation signal, then makes comparison about computation burden between FFT and FFS. Simulation result shows that the algorithm proposed by this paper has better reconstruct precision. And when signal length increases sharply, its computation is influenced not much. So its run time for large amount of signal is far smaller than FFT.
出处
《自动化技术与应用》
2016年第4期30-35,共6页
Techniques of Automation and Applications
关键词
快速傅里叶采样
压缩感知
认知无线电
频谱感知
fast fourier fampling
compressed sensing
cognitive radio
spectrum sensing
