We present a novel approach to cooperative aerial transportation through a team of drones, using optimal control theory and a hierarchical control strategy. We assume the drones are connected to the payload through rigid attachments, essentially transforming the whole system into a larger flying object with "thrust modules" at the attachment locations of the drones. We investigate the optimal arrangement of the thrust modules around the payload, so that the resulting system is robust to disturbances. We choose the $\mathcal{H}_2$ norm as a measure of robustness, and propose an iterative optimization routine to compute the optimal layout of the vehicles around the object. We experimentally validate our approach using four drones and comparing the disturbance rejection performances achieved by two different layouts (the optimal one and a sub-optimal one), and observe that the results match our predictions.