%0 Journal Article %T 一种基于正交匹配追踪的水下目标相关检测方法<br>Signal Spatial Correlation Improvement and Correlation Detection Method Based on Sparse Reconstruction %A 邵炫 %A 孙 %A 超 %A 孔德智 %A 蒋光禹 %J 西北工业大学学报 %D 2017 %X 相关检测是一种常见的水下目标检测方法,其性能易受到空间相关性衰减的影响。针对水下信号空间相关性衰减导致常规相关检测方法性能下降的现象,提出了一种基于正交匹配追踪的水下目标相关检测方法。该方法利用接收信号在字典原子上的稀疏表示特性,利用正交匹配追踪方法从过完备字典原子集合中选取尽可能少的字典原子来恢复接收信号;在恢复过程的同时通过设定阈值保留信号的本质特征并舍去易受环境扰动和噪声影响的小特征成分,起到改善接收信号的空间相关性,提高空间相关性衰减情况下相关检测方法性能的作用。研究表明,稀疏重构可以消除环境扰动和噪声干扰的影响,有效地恢复信号的空间相关特性;该方法检测性能明显优于传统相关检测方法。利用蒙特卡罗实验对所提检测方法的有效性进行仿真验证。<br>Correlation detection method is usually applied to underwater target detection, the detection performance of which is affected by spatial correlation attenuation easily. In the paper, a correlation detection method based on sparse reconstruction via orthogonal matching pursuit is proposed. Based on the sparse representation property of the received signal, the method recovers the received signal by selecting the smallest number of dictionary atoms from the set of over-complete dictionary atoms by using the orthogonal matching pursuit method. In the process of the signal reconstruction, the essential characteristic components of the signal are retained and the small characteristic components affected easily by environmental disturbance and the noise are abandoned through threshold setting. Results show that sparse reconstruction can reduce environmental effects and improve the spatial correlation characteristic, and that the correlation detection method based on sparse reconstruction performs better than the traditional correlation detection method. The results are validated by computer simulations %K 空间相关性 %K 正交匹配追踪 %K 压缩感知 %K 稀疏重构 %K 相关检测 %K 水下声学 %K 最大似然 %K 蒙特卡罗方法 %K 信号处理 %K 信噪比< %K br> %K signal spatial correlation %K orthogonal matching pursuit %K compressed sensing %K sparse reconstruction %K correlation detection %K underwater acoustics %K maximum likelihood %K Monte Carlo methods %K signal processing %K signal to noise ratio %U http://journals.nwpu.edu.cn/xbgydxxb/CN/abstract/abstract6809.shtml