Partitioning logic functions for minimum size implementations

When implementing large logic functions there is a need to partition the function to fit the implementation primitives. There is, at present, no simple analytic method of achieving a `best´ implementation. In this paper, classical decomposition theory is extended to allow the analysis of functions with large input fields and multiple outputs, specified by their ON-terms and including `don´t-care´ input values. A method, based on intersecting incompatible tuples is developed to determine the best decomposition of such complex logic functions