The general minimum-mean-square error (MMSE) nonlinear filter is tractable if it is restricted or reduced to have no memory. The importance of the zero-memory nonlinear (ZNL) filter is that it is a fundamental building block for more general nonlinear systems. This paper discusses several problems related to ZNL filtering, in particular, one aspect of the prefiltering problem. The general solution of an inverse problem to the prefiltering problem is discussed in detail, and the cases where it assumes a comparatively simple form are treated. Two examples are given to illustrate the characteristics of the nonlinearities and the resulting errors.
ASJC Scopus subject areas
- Control and Systems Engineering
- Signal Processing
- Computer Networks and Communications
- Applied Mathematics