In this contribution we consider the problem of optimal drone positioning for improving the operation of a mobile ad hoc network. We build upon our previous results devoted to the application of game-theoretic methods for computing optimal strategies. One specific problem that arises in this context is that the optimal solution cannot be uniquely determined. In this case, one has to use some other criteria to choose the best (in some sense) of all optimal solutions. It is argued that centrality measures as well as node ranking can provide a good criterion for the selection of a unique solution. We showed that for two specific networks most criteria yielded the same solution, thus demonstrating good coherence in their predictions.