基于稀疏分解的多分量线性调频信号参数估计  

Multicomponent chirp signal parameter estimation via sparse representation

作　　者：郭金库[1,2] 吴瑾颖[2] 张茜[2] 

机构地区：[1]清华大学自动化系信息所,北京100084 [2]第二炮兵工程学院,西安710025

出　　处：《振动与冲击》2011年第1期175-179,共5页Journal of Vibration and Shock

基　　金：国家自然科学基金项目(60672108);国家自然科学基金项目(60372020)资助

摘　　要：提出了一种多分量线性调频信号的参数估计方法。基于过完备Gabor字典的Matching Pursuit算法,可以将信号表示为Gabor原子的线性组合。这些原子有效的揭示了信号的内在时频结构特征,是信号的一种稀疏表示。该方法直接利用分解得到的稀疏信息对信号中调频分量的调频率、初始频率和结束频率进行估计。仿真结果显示,该方法适用于存在强有意干扰或者有色噪声的环境。Here, a novel algorithm for parameter estimation of multicomponent chirp signal in complicated noise environment was proposed. Using the matching pursuit algorithm, the signal was firstly decomposed into Gabor atoms which provided sparse information representing the signal time-frequency feature. Hough transformation was then directly used to estimate chirp rate, starting frequency and stopping frequency of chirp components without computing the time- frequency distribution. Simulation results showed that this algorithm is capable of estimating parameters of muhicomponent chirp signal, even in the presence of powerful intended interference and colored noise ; comparing to other techniques, the algorithm is computationally efficient and possible to be used in real-life problems.

关 键 词：稀疏分解 线性调频信号 MATCHING Pursuit算法 Gabor字典 HOUGH变换 

分 类 号：TN911.72[电子电信—通信与信息系统]