Multilinear Karhunen-Loève Transforms

• Published in: IEEE transactions on signal processing 2022, Vol.70, p.5148-5163
• Main Authors: Howlett, Phil, Torokhti, Anatoli, Pudney, Peter, Soto-Quiros, Pablo
• Format: Article
• Language: English
• Subjects: Computing costs; Covariance matrices; Estimation; Image reconstruction; Karhunen-Loève transform; Loss measurement; Matrix decomposition; Optimization; Principal component analysis; Reconstruction; Signal processing; Transforms; Transmitters; Wiener filtering
• ISSN: 1053-587X; 1941-0476
• DOI: 10.1109/TSP.2022.3214684

Many recent applications involve distributed signal processing where a source signal is observed by, say, p local receiver-transmitters and then transmitted to a reconstruction center for the signal estimation. An optimal determination of the receiver-transmitters and the reconstruction center requires extensions of the Karhunen-Loève transform (KLT) and Wiener filter. In this paper, the associated extensions are provided. The proposed optimal multilinear filter is a generalization of the Wiener filter and consists of p terms where each term is associated with a local receiver-transmitter. For the case when the receiver-transmitters must reduce the dimensionality of the observed signals, two associated techniques are proposed: the multilinear KLT-1 and multilinear KLT-2. The multilinear KLT-1 is constructed in terms of pseudo-inverse matrices and therefore always exists. The multilinear KLT-2 is given in terms of non-singular matrices and it may provide a higher associated accuracy than that of the multilinear KLT-1. All three proposed techniques are based on a reduction of the original problem to p separate error minimization problems with small matrices. This allows us to provide a fast computational procedure for the multilinear filter, and decrease the computational cost for constructing the multilinear KLT-1 and KLT-2.