蒸馏稀释感知:适应性采样下的稀疏检测与估计

Distilled-diluted sensing:Adaptive sampling for sparse detection and estimation

针对稀疏信号采样问题,提出一种改进的适应性采样方法,通过对原始信号的多次观测逐渐筛选出有用信号,从而提高观测的准确度.该方法增加了一个找回误判分量的机制,弥补了原始方法无法抑制漏检率的缺点.实验结果表明,所提出的方法有效地降低了判决器的漏检率,同时保持相同的计算复杂度,且在中等及较小信噪比条件下性能更优.

For the sparse signal sampling problem, an improved adaptive sampling method is proposed. The proposed method offsets the drawback of distilled sensing of high non-detection rate. The simulation results show that the proposed method effectively decreases the estimator＇s non-detection rate with commensurate complexity, and show great performance in the middle and relatively low SNR condition.