Advancing evolutionary coordination for fixed-wing communications UAVs

In this paper we present advances to our previously proposed coordination system for groups of unmanned aerial vehicles that provide a network backbone over mobile ground-based vehicles. Evolutionary algorithms are employed in order to evolve flying manoeuvres that position the aerial vehicles. The updates to the system include obstacle representation, a packing mechanism to permit efficient dynamic allocation of ground-based vehicles to their supporting aerial vehicles within large-scale environments, and changes to time synchronisation. The experimental results presented in this paper show that the system is able to adaptively form sparse formations that cover as many ground-based vehicles as possible, optimising the use of the available power.