Comparative studies on the performance of heuristic approaches such as PSO, GA and ACO in urban and highway scenarios show an improvement in all performance metrics except bandwidth. Also the results of these three tools are quite close.
For more details and information about this subject, i suggest you to see links on topic.
An optimized routing algorithm for vehicle ad-hoc networks ...
Referring to the No Free Lunch (NFL) Theorem, "any two optimization algorithms are equivalent when their performance is averaged across all possible problems". Therefore, no one can claim that Algorithm X, which is a general meta-heuristic one, is the best for optimizing a given problem. One should adapt and customize it to exploit its advantages as much as possible.
Since Ant Colony Optimization (ACO) Algorithm is intrinsically designed to solve routing problems, I think it can be in your special case, i.e. Vehicular ad-hoc network routing, be more useful if you tune its parameters properly. It should be noted that if your network grow unreasonably, updating ACO's tables takes too much time and applying this algorithm is not logical.
Computational speed, solution quality, and programming complexity all factor into which one is chosen. Tests on standard problems may reveal an answer, but the choice is always difficult. Simple s-metaheuristics can work as well as p-metaheuristics in these types of problems.