Abstract
Extremal convex and concave extensions of Boolean functions—used in solving and analyzing systems of Boolean equations, a fundamental problem in computer science—can be constructed based on a set of weighting coefficients that allow a point x in an n-dimensional space to be represented as a convex combination of the vertices of an n-dimensional unit cube. In this paper, the properties of such a set of weighting coefficients and one partition of the n-dimensional unit cube are substantiated. Based on these, an explicit analytical (simplified) solution is derived for the problem of calculating the value of a concave extension of a Boolean function, which is a particular problem of linear programming, for each of the four particular cases corresponding to these properties. Specifically, it is proven that if the set of weighting coefficients consists of a single element, then the values of the minimal concave extension and the polylinear extension of an arbitrary n-ary Boolean function coincide. Explicit formulas are provided for the minimal concave extensions of two fundamental n-ary Boolean functions: the disjunction and the conjunction of rank n. Furthermore, an explicit formula is derived for the minimal concave extension of an n-ary Boolean function whose zeros are non-adjacent. This work complements the existing literature on the properties of sets of weighting coefficients and the extremal convex and concave extensions of Boolean functions.