Riemannian mean and space-time adaptive processing using projection and inversion algorithms
Lists
Only you see this
Abstract
The estimation of the covariance matrix from real data is required in the application of space-time adaptive processing (STAP) to an airborne ground moving target indication (GMTI) radar. A natural approach to estimation of the covariance matrix that is based on the information geometry has been proposed. In this paper, the output of the Riemannian mean is used in inversion and projection algorithms. It is found that the projection class of algorithms can yield very significant gains, even when the gains due to inversion-based algorithms are marginal over standard algorithms. The performance of the projection class of algorithms does not appear to be overly sensitive to the projected subspace dimension.
BibTeX
@inproceedings{balaji2013riemannian,
author = {Bhashyam Balaji and Frédéric Barbaresco},
title = {{Riemannian mean and space-time adaptive processing using projection and inversion algorithms}},
booktitle = {Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE},
year = {2013},
doi = {10.1117/12.2017878},
}Loading related content...