超宽带信号的快速匹配追踪算法

Fast Matching Tracking Algorithm for UWB Signals

压缩感知理论可应用到脉冲超宽带通信中,以降低采样速率。超宽带信号常常采用多径分集字典来稀疏表示,使得压缩采样匹配追踪(CoSaMP)、广义正交匹配追踪(GOMP)等经典的重构算法经常出现原子误选的情况,导致信号重构误差增大。因此,在GOMP的基础上,提出了一种新的重构算法。该算法在每次迭代中可选取多个原子以加快迭代收敛速度,并且通过限制被选中原子之间的序号差来避免原子误选。仿真实验表明,该算法不仅可以保证重构精度,而且具有更低的计算复杂度。

Compressive sensing theory may be applied to Ultra-Wideband Impulse Radio communications, so as to reduce the sampling rate. Ultra-wideband signals are usually sparsely represented by multipath diversity dictionaries, and this makes the classical reconstruction algorithms such as ComsaMP, GOMP （Generalized Orthogonal Matching Pursuit） and other classic reconfiguration algorithms often suffer from incorrect selection of atoms, leading to the increase of signal-reconstruction error. Therefore, based on GOMP, a new reconstruction algorithm is proposed. This algorithm may select multiple atoms in each iteration and thus speed up the convergence of iterations, and avoid atom misselection by limiting the sequence number difference of between the selected atoms. Simulation experiments indicate that this algorithm could guarantee the accuracy of reconstruction and has fairly low computational complexity.