摘要
现有的基于特征分解类的角度超分辨算法在理想阵列条件下,其估计性能良好,但当信号模型与实际信号环境不匹配,即存在系统误差时,算法的估计性能会严重下降,甚至失效。针对此问题,提出了一种波达方向估计和通道误差校正的新算法,先对各子阵接收信号预处理,使得通道误差对每个子阵的影响一致,这样阵列旋转因子就与通道误差无关,再利用阵列旋转不变性,实现波达方向估计和阵列通道误差的校正。无需校正源,也不需要知道通道误差的先验信息,无需多维搜索寻优和迭代。理论分析和计算机仿真都表明新算法的优越性。
Many high resolution direction-of-arrival（DOA）estimation methods based on eigenvalue decomposition have excellent estimation performance under ideal condition, but their performance degrades significantly and even can not work when the model does not match with the real condition due to system errors. To deal with this problem, an algorithm of DOA and array calibration is proposed, the influences of array gain and phase errors to every suharray are all same by preprocessing to received data, that is said the rotation factor is no relation with array gain and phase errors, so the DOA estimation and array calibration are implemented by using ro rational invariance techniques. The method doesn＇t need reference signal source, a prior knowledge of array gain and phase errors, multidimensional search, optimization and iteration. Theoretical analysis and simulation results demonstrate the merits of the new algorithm.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2008年第10期1847-1850,共4页
Systems Engineering and Electronics
关键词
波达方向
阵列通道误差
旋转不变性
阵列校正
direction-of-arrival
array channel error
rotational invarianee technique
array calibration
