作者: M. Viberg , A.L. Swindlehurst
DOI: 10.1109/78.330367
关键词: Subspace topology 、 Asymptotic analysis 、 Covariance 、 Computer science 、 Estimation theory 、 Signal processing 、 Array processing 、 Statistics 、 Algorithm 、 Sensor array 、 Antenna array 、 Noise (signal processing) 、 Weighting
摘要: The principal sources of estimation error in sensor array signal processing applications are the finite sample effects additive noise and imprecise models for antenna spatial statistics. While these errors have been studied individually, their combined effect has not yet rigorously analyzed. authors undertake such an analysis class so-called subspace fitting algorithms. In addition to deriving first-order asymptotic expressions error, they show that overall optimal weighting exists a particular covariance model. companion paper, optimally weighted method is shown be asymptotically equivalent with more complicated maximum posteriori estimator. Thus, model question, no other can yield accurate estimates large samples small errors. Numerical examples computer simulations included illustrate obtained results verify realistic scenarios. >