Deconvolution of linear systems by constrained regression and its relationship to the Wiener theory

In this paper we discuss the problem of deconvolution of the output of a linear system in the presence of noise. As formulated the problem is solvable by least-squares linear regression. A previously known technique for solving integral equations is applied. It is shown now this solution is equivalent to constrained linear regression and that this may be computed in the frequency domain. Finally, the relationship between deconvolution by constrained linear regression and by Wiener theory is derived.