Air Traffic Management Using a GPU-Accelerated Genetic Algorithm

Abstract

Air traffic management is becoming highly complex with the rapid increase in the number of commercial and cargo flights, leading to increased traffic congestion and flight delays. To mitigate these issues, we present a flight path generation system that distributes the aeroplanes across the airspace and imparts minimal delays to the flight if required, thus ensuring that the aircraft follows the shortest route wherein it encounters the least amount of traffic. We develop a parallel genetic algorithm in CUDA-C with a novel fitness function allowing the system to reach an optimal solution where the air traffic density is minimised. The proposed algorithm was tested on one day's domestic flight schedule and achieved an 18% reduction in traffic density, with the flight times and delays remaining proportional to the data observed in the existing air traffic management system.