In the companion paper, a single-level mathematical formulation has be
en presented describing the multiple campaign planning/scheduling prob
lem in considerable generality. However, the resulting mixed integer l
inear programming model is too large to be computationally tractable f
or many cases of practical interest. In this paper, we present a rigor
ous decomposition approach to the solution of this problem and demonst
rate its effectiveness by applying it to a number of illustrative exam
ples. In addition, we consider ways in which the structure of the cons
tituent mathematical models of the decomposition scheme can be exploit
ed to reduce their sizes and the associated integrality gaps. Examples
illustrating the applicability of the overall approach are also prese
nted.