中文摘要：

正交压缩采样是低速获取带通模拟信号同相和正交分量的新型模信转换系统,可广泛应用于雷达、通信等电子系统。但是对于宽带或超宽带脉冲雷达,重构奈奎斯特率的全程回波信号需要大的存储空间和计算量,以致于难以实现实时重构。该文在对正交压缩采样系统特性进行分析的基础上,将测量矩阵近似成一种具有特殊带状结构的矩阵,然后采用分段滑动重构思想实现实时重构。仿真结果表明,在对测量矩阵进行合理近似的基础上,该文提出的重构方法可以极大地节省存储空间和计算时间,实现近似最优的重构性能。

英文摘要：

Quadrature Compressive Sampling（Quad CS） is an efficient Analog-to-Information Conversion（AIC） system to sample band-pass analog signals at sub-Nyquist rates. The Quad CS can be widely used in radar and communication systems to acquire sub-Nyquist samples of inphase and quadrature components. However, for wideband or ultra-wideband pulsed radars, it is often impractical to reconstruct Nyquist samples of full-range echoes in real-time because of huge storage and computational loads. Based on the characteristics of Quad CS system, an approximate scheme is proposed to transform the Quad CS measurement matrix into a matrix with a special banded structure. With the banded matrix, a segment-sliding reconstruction method is adopted to perform real-time reconstruction. Simulation results show that with a reasonable approximation of the measurement matrix, the proposed reconstruction scheme achieves nearly optimal reconstruction performance with a significant reduction of data storage and computational time.