Evolving potential field parameters for deploying UAV-based two-hop wireless mesh networks

ABSTRACTWe attack the problem of dynamic UAV-based ad-hoc mesh network deployment to find and connect people in an area of interest. Genetic algorithms tune the parameters of potential fields that control UAV movement in order to optimize people coverage and network longevity. We extend earlier work that assumed one-hop communication between UAVs to the more realistic two-hop case and find significant increase in coverage and network longevity. Experimental results show that enabling two-hop communications between UAVs improves the performance on average between 4% to 32.19%.