Monotone iterative methods for a general class of discrete boundary value problem (Q1334588)

The authors study discrete boundary value problems of the following type \[ (P)\;x_{k+1} = f_ k(x_ 0,x_ 1,\ldots,x_ k), \qquad Ax_ 0+Bx_ N=c, \] where \(A\) and \(B\) are given matrices, \(c\) a given vector, and \(x_ k=(x^ 1,x^ 2,\ldots,x^ n)(t_ k)\) is a mapping from the discrete set \(T=(t_ 0,t_ 1,\ldots,t_ N)\) into \({\mathbb{R}}^ n\). This type of problem includes, for example, finite as well as infinite delay equations and (discrete) integral equations. In the first main part of the paper, comparison results are given for those problems of the above-mentioned type, which satisfy a certain additional monotonicity condition (``Condition \(C(p,q)\)''). Furthermore, essentially two different monotone iterative schemes for the computation of quasi-solutions of the problem \((P)\) are developed. A number of numerical examples illustrates quite nicely the authors' methods and results.