Divided Difference Methods for Galois Switching Functions

An alternative is provided to a recently published method of Benjauthrit and Reed for calculating the coefficients of the polynomial expansion of a given function. The method herein is an adaptation to finite fields of a method of Newton. The method is exhibited for functions of one and two variables. The relative advantages and disadvantages of the two methods are discussed. Some empirical results are given for GF(9) and GF(16). It is shown that functions with "don't care" states are represented by a polynomial of minimal degree by this method.