Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
We derive an efficient recursive procedure for the triangular factorization of strongly regular matrices with generalized displacement structure that includes, as special cases, a variety of previously studied classes such as Toeplitz-like and Hankel-like matrices. The derivation is based on combining a simple Gaussian elimination procedure with displacement structure, and leads to a transmission-like interpretation in terms of two cascades of first-order sections. We further derive state-space realizations for each section and for the entire cascades, and show that these realizations satisfy a generalized embedding result and a generalized notion of J-losslessness. The cascades turn out to have intrinsic blocking properties, which can be shown to be equivalent to interpolation constrains.
Cécile Hébert, Duncan Alexander, Nathanaël Perraudin, Hui Chen
Daniel Kressner, Meiyue Shao, Yuxin Ma