Upon just looking quickly on the graphs, it appears to not be TAP or QAP, but more like a problem over a tree with perhaps some extra links. The information is not really ideal to get well informed. I suggest you find a scholar in that exact field, and get real science out of it.
The ACO meta heuristic has been defined by their authors (Dorigo et al.) when they try to generalize their Ant System heuristic used first for the TSP for other kinds of problems, under four conditions:
To see the problem as a search graph;
To have an auto-catalytic process (a positive reinforcement);
To have an Heuristics which allows a constructive definition of the solution (the greedy force);
To have a constraint satisfaction method (stopping criteria).
The ACO meta heuristic can be and have been used for QAP and WSN (both QAP and WSN's applications verify the four previous conditions).
Your confusion is, so, very obvious and inevitable :
You try to apply the ACO meta heuristic for WSN's applications using one of two combinatorial optimization problems: TSP ( Travelling Salesman Problem) or QAP (Quadratic Assignment Problem) !!!!!!!!!!!!!!!!!!!!!!!!!!!
Applying a meta-heuristic for a problem using a problem!
I'm missing some thing or I miss-understand your problem?