Using dispersion-induced group delay to solve the integer ambiguity problem: a theoretical analysis