The maximum principle and growth estimates for partial difference equations (Q2490645)

This paper deals with the linear partial difference equation \(\sum_{\beta\in B}a(\alpha, \beta)y(\alpha+\beta)=x(\alpha)\), \(\alpha\in A\), where \(A\subset {\mathbb Z}^n\), \(B\subset {\mathbb Z}^n\), \(1<| B| <+\infty\). The domain \(A\) will be an infinite set and \(x\) is a mapping \(x: A\to {\mathbb C}\). By means of the existence and uniqueness theorem for initial value problems, the author obtains the maximum principle: under certain conditions the values of the solutions are bounded by the norm of the initial values (in the homogeneous case), or bounded by the norm of the input (in the non-homogeneous case with zero initial values). Theses conditions, demanding that the absolute value of a certain coefficient is dominant, i.e. greater than or equal to the sum of absolute values of the remaining coefficients, are easy to verify. Such a maximum principle is further applied to obtain various growth estimates. Examples include growth estimates for Stirling numbers, Euler numbers, coefficients of orthogonal polynomials and other sequences defined by partial difference equations. Generalizations for systems of partial difference equations and some nonlinear partial difference equations are also formulated.