Micro Unmanned Aerial Vehicles (MUAV) emerge more and more to universal platforms which can be equipped with a variety of sensors. The use of cognitive MUAV swarms in applications for the distribution of robots and mobile sensors is an appreciable benefit for search-and-rescue and surveillance missions. In this paper we particularly focus on an agent-based methodology for communication aware motion behavior of self-organizing MUAVs in remote sensing missions. The objective of our proposed cognitive and sensor aided mesh network is to maximize the spatial sensing coverage on the one hand and the connectivity between the MUAVs on the other hand. Next to these optimization goals we developed algorithms seeking concurrently for a global target and a coherence of the swarm in order to avoid self-separations. To determine these key figures, the performance of different proposed algorithms is analyzed and compared.