This thesis investigates new schemes to improve the operability of heterogeneous Unmanned Aircraft Systems (UAS) teams through the exploitation of inter-vehicular communications. Releasing ground links from unnecessary data exchanges saves resources (power, bandwidth, etc) and alleviates the inherent scalability problem resulting from the increase in the number of UAS to be controlled simultaneously. In first place, a framework to classify UAS according to their level of autonomy is presented along with efficient methodologies to assess the autonomy level of either individual or multiple UAS. An architecture based on an aerial Mobile Ad-hoc Network (MANET) is proposed for the management of the data exchange among all the vehicles in the team. A performance evaluation of the two most relevant MANET approaches for path discovery (namely, reactive and proactive) has been carried out by means of simulation of two well-known routing protocols: Ad-hoc On-demand Distance Vector (AODV) and Destination Sequenced Distance Vector (DSDV). Several network configurations are generated to emulate different possible contingencies that might occur in real UAS team operations. Network topology evolution, vehicle flight dynamics and data traffic patterns are considered as input parameters to the simulation model. The analysis of the system behaviour for each possible network configuration is used to evaluate the appropriateness of both approaches in different mission scenarios. Alternative network solutions based on Delay Tolerant Networking (DTN) for situations of intermittent connectivity and network partitioning are outlined. Finally, an assessment of the simulation results is presented along with a discussion about further research challenges.