The analysis of structurally constrained minimum mean-square error block matrix filters

We consider here the problem of applying system requirements in SISO FIR block matrix filters, their least-squares error formulation with a look towards improving computational efficiency, and their eventual performance characterization. It is shown that the proposed least squares formulation can be applied to structural requirements in general and that the approach taken in this paper is adequate for estimation and detection applications with the ultimate goal of real-time processing.