Least-squares wave front errors of minimum norm

Requirements for adaptive optics and compensating imaging systems lead to wave front reconstruction problems which we formulate as generalized least-squares problems. For a given array of phase-difference measurements, we construct explicit and exact solutions for the least-squares wave front error. Of particular interest are solutions with minimum norm. Two different discretizations for the gradient are used and the reasons for the different results are given.