基于再采样和Toeplitz重构的改进信源数目估计方法  

Enhanced Source Number Estimator Based on Resampling and Toeplitz Reconstruction

作　　者：江横 李立春[1] 蒲敏刚 张海龙[1] JIANG Heng;LI Lichun;PU Mingang;ZHANG Hailong(Information Engineering University,Zhengzhou 450001,China)

机构地区：[1]信息工程大学,河南郑州450001

出　　处：《信息工程大学学报》2022年第3期291-295,共5页Journal of Information Engineering University

摘　　要：为解决现有信源数目估计方法在低快拍数下准确率较低的问题,引入再采样方法充分利用有限的接收数据信息以交叉验证估计结果,提高估计的准确率。针对超定和欠定条件下分别给出两种策略:超定条件下,直接应用再采样优化现有的信源估计方法;欠定条件下,结合稀疏阵型使用Toeplitz重构法,增加阵列自由度和特征值区分度,并将再采样方法与改进的滑动窗法结合来优化信源数目估计。理论分析和数值仿真验证所提方法的优越性。To solve the dilemma that the existing source number estimator possesses low accuracy under the scenario of small snapshots,the resampling method is introduced to make full use of the limited received data information to cross-verify the estimation results and improve the accuracy.Two strategies for source number estimation under overdetermined and underdetermined conditions are presented respectively:under overdetermined conditions,the resampling theory is directly applied to optimize the existing source number estimators;in underdetermined conditions,Toeplitz reconstruction is exploited based on sparse array to increase the degree of freedom and eigenvalue discrimination.Further,the resampling method is combined with modified sliding window scheme to estimate the number of sources.The superiority of the proposed method is validated by theoretical analysis and numerical simulation.

关 键 词：信源数估计 再采样 Toeplitz重构 

分 类 号：TN911.7[电子电信—通信与信息系统]