Comparative representations of a genetic algorithm to locate unmanned aerial vehicles in disaster zones

Our economy and society depend on the continuous operation of the internet and other wireless networks. However, during or after a natural disaster, communications infrastructure can be affected and even interrupted. Effective planning of emergency operations in these scenarios can play an essential role in saving lives. Recently, the use of Unmanned Aerial Vehicles (UAVs) has been proposed to provide broadband connectivity. UAVs can be rapidly deployed as aerial base-stations over the affected area and provide connectivity between victims and emergency operators. However, one of the challenges for their deployment in emergency scenarios is finding their optimal locations to provide the largest number of communication services. This paper introduces an optimization model which positions UAVs in such a way as to maximize their coverage (the number of mobile users covered), thus guaranteeing a successful voice service in an LTE network. A genetic algorithm (GA) with a steady-state population configuration is used to find optimal locations of the UAVs. We present the results of the GA using two different representations: binary and floating-point. The results indicate that the genetic algorithm with a steady-state model performs better using a binary representation.