Drone Trajectory Optimization using Genetic Algorithm with Prioritized Base Stations

Drones have been widely applied to perform emergent tasks in the post-disaster scenario, due to their unique characteristics such as mobility, flexibility, and adaptivity to altitude. However, drones have limited energy capacity, which presents a major drawback in flight time and affects their performance in such scenarios. Hence, trajectory optimization has become a critical research problem for such applications of drones. In this paper, we present an optimal trajectory design for a single drone to ferry data from temporary Base Stations (BSs) deployed within a disaster zone to a fixed gateway node that is the point of origin and final destination for the drone flight. We have used a Genetic Algorithm (GA)-based approach that takes into account the shortest distance traveled and least time spent by the drone during flight. We also examine the case where some BSs have delay requirements that are unknown to the drone in advance. Simulation results show that the performance of our proposed GA-based approach matches that of the benchmark exhaustive search algorithm and the difference in computational time between the 2 algorithms increases with the number of BSs. Accordingly, our proposed algorithm has 96.4% lower computational time complexity compared to the benchmark exhaustive search algorithm when there are 12 BSs in the disaster area.