Least squares solutions in signal processing using the singular value decomposition
John Mather · Defense Technical Information Center (DTIC) · 1986
We compare a number of advanced antenna array and time series processing techniques which either utilise singular value decomposition, or may be profitably interpreted in terms of it. Computer simulations of these techniques acting on spatio-temporal data from linear, conformal circular and focal plane arrays are presented and discussed. Means of introducing prior knowledge are described, and lead to the suggestion of a computationally efficient processing approach. We have made extensive use of a sequential processing strategy for multiple domain data and show by example that this potentially allows both detection and classification of additional emitters.