Dynamic airspace configuration through genetic algorithm optimization using air traffic complexity

Dynamic Airspace Configuration (DAC) aims to reconfigure control-sector boundaries in real time so that air-traffic-controller (ATCO) workload remains balanced even under fluctuating demand. This paper proposes and validates an end-to-end DAC optimization module that relies on a complexity-driven Genetic Algorithm (GA), an optimal layer-based 3-D slicing technique, and a post-processing step that converts discrete polygonal sectors into continuous, operationally feasible geometries. The workflow begins by mapping complexity onto a 20 × 20 nm grid that preserves the baseline CNF5A sectorization. A Simple GA swaps adjacent cell pairs, minimizing the variance and max–min gap of sector complexity while enforcing strict connectivity constraints. Tests on Madrid North ACC traffic cut the five-sector complexity gap from 10.34 to 0.81 and the four-sector gap from 6.32 to 0.78. A vertical layer-cut routine finds the best upper/lower split near FL365; running the GA above and below this level and smoothing boundaries yields a balanced nine-sector 3-D layout ready for operational trials.