An Iterative Wiener Filter for the Identification of Multilinear Forms
Laura-Maria Dogariu, Silviu Ciochină, Constantin Paleologu, Jacob Benesty, Cristina C. Oprea · 2020
There are many important fields nowadays which involve multilinear system identification. In this context, the number of parameters of the global impulse response to be identified can represent an important challenge, leading to the need for tensorial decomposition and modelling of such systems. We recently focused on the identification of particular types of multidimensional problems, such as bilinear and trilinear forms. In this paper, our purpose is to extend the solutions obtained for such particular cases to the general case of higher-order multilinear systems. To this aim, we propose an iterative Wiener filter tailored for the identification of multilinear forms of any order. Simulations support the theoretical findings and prove the good performance of the proposed solution.