Iterative methods for the extremal positive definite solution of the matrix equation \(X+A^{*}X^{-\alpha}A=Q\) (Q869490)

The authors propose two algorithms that avoid matrix inversion for every iteration, called inversion free variant of the basic point iteration. The first algorithm computes the maximal positive definite solution \(X_{+}\) of the nonlinear equation \(X+A^{\ast }X^{-\alpha }A=Q,\) where \(A\) is \ a nonsingular matrix, \(Q\) is a Hermitian positive definite matrix and \(\alpha \in (0,1].\) The second algorithm computes the minimal positive definite solution \(X_{-}\) of the same nonlinear equation with the case \(\alpha \in [ 1,\infty )\). Convergence theorems are provided. Some numerical examples are added to illustrate the convergence features.