基于迭代自适应稀疏分解的雷达信号去噪

稀疏分解能有效分离信号和噪声,因此适用于信号去噪.文中构造了雷达回波稀疏表示的冗余字典,字典原子与目标回波波形匹配,基于该字典的雷达回波信号稀疏度就是目标数.针对稀疏度自适应匹配追踪算法进行低信噪比信号稀疏分解时的不足,提出了一种迭代自适应匹配追踪算法,采用规范化的残差之差作为迭代终止条件,使得稀疏分解过程能依据噪声水平自适应终止,以逐次逼近方式估计信号稀疏度,改善了稀疏分解的精度.仿真实验结果表明,该算法在低信噪比以及稀疏度未知的条件下,实现了雷达回波信号的准确稀疏分解,极大地提高了信噪比.

Radar Signal Denoising Via Adaptive Iterative Sparsity Decomposition

Sparse decomposition is effective in separating signal and noise, and it can be used to remove noise. In this paper, a redundancy match dictionary is designed for radar echo signal sparse representation, and the signal sparsity is equal to the detecting target number. As the stop threshold of the sparsity adaptive matching pursuit(SAMP) algorithm is not applicable for sparse decomposition in low signal to noise ratio(SNR) conditions, the iteration adaptive matching pursuit(IAMP) algorithm is proposed using normalized residual difference as stop condition, making sparse decomposition adaptively stop according to noise level. Signal sparsity estimation is implemented by way of successive approximation, and much improvement on decomposition accuracy is obtained. Extensive simulation results show that the IAMP algorithm is effective in radar echo signal sparse decomposition in low SNR conditions without sparsity information, and the SNR of sparse decomposition signal is largely improved.