The addition to the INETMANET framework of OMNeT++ uses cross-layer approach to utilize application layer knowledge about controlled mobility to enhance the routing decisions in mobile robotic networks. We use the well-known mobility model from Reynolds as a basic structure for our logical submodules, which handle specific tasks like exploration, collision avoidance and maintenance of the swarm coherence. Different novel routing metrics use the results of a new prediction method for future trajectories for their decision-making.