System’s
least-squares inversion
•For computing the matrix of N filtering coefficients hik, a least-squares method is employed.
•A “total squared error” etot is
defined as:
•A set of N linear equations is formed by minimising etot, imposing that:
