Domination problems with no conflicts

In recent works, authors added conflicts to classical discrete optimization problems. In this paper, a conflict is a pair of vertices that cannot be both in a solution. Set of conflicts can be viewed as edges of a so called conflict graph. An instance is then a support graph and a conflict graph. With these new constraints, the existence of a solution (dominating set or independent dominating set) with no conflicts is no more guaranteed. We explore this subject and we prove that it is NP-complete to decide the existence of a solution even in very restricted classes of graphs and conflicts (sparse or dense). We also propose polynomial algorithms for some sub-cases, using deterministic finite automata.