模拟压缩采样技术研究

Analog compressed sampling technology

高带宽信号的应用对传统的香农-奈奎斯特采样提出了苛刻挑战,因此,基于压缩感知理论,对模拟信号的亚奈奎斯特采样技术进行研究。将压缩感知概念从离散域推广到模拟信号域,采用压缩采样算子对模拟信号输入带宽进行压缩,实现模拟压缩采样。在模拟压缩采样算子设计中,基于信号结构选择能够与信号稀疏有效结合成单一硬件结构的感知矩阵,使在压缩带宽的同时不损失任何重要信息。在压缩采样的数字处理中,应用非线性探测方法从低速率压缩采样序列中探测信号的精确子空间。

A variety of applications of signals with large bandwidth have imposed increasing rigorous demanding challenges on the traditional Shannon-Nyquist sampling method. Based on compressed sensing theory, the technology of sub-Nyquist sampling of analog signals is discussed in the paper. With the extension of compressed sensing ideas to analog domain, analog compressed sampling operator narrows down the analog input bandwidth. The design of analog compression operator should first consider the input signal structure model, and then chose an appropriate sensing matrix which can be combined with the signal sparse to a single hardware, in order that no import information is lost when analog band is compressed. For the digital processsing of the compressed samples, using no-linear method, the accurate signal subspce can be found from the sampled sequences at low rate.