Dynamic Airspace Configuration Management Based on Computational Geometry Techniques

Abstract

New techniques for dynamic airspace configuration in the National Airspace System based on computational geometry techniques are investigated. The current airspace is subdivided into Air Route Traffic Control Centers and sectors of airspace that remain statically defined. New methods of designing airspace regions that are balanced in terms of workload per unit area over a given planning time period are presented. Configuring the airspace dynamically provides a means to change sector designs from one day to the next or within the course of a single day, to better balance controller workload. A recursive, top-down partitioning algorithm is used to subdivide a given 2D polygonal region R of airspace (e.g., a center) into sector regions. Three types of local partitioning are investigated: straight-line cuts, pie-cuts, and wheel-cuts. During each subdivision, the local partitioning balances workload among the resulting subregions, while other shape parameters, such as aspect ratio, are kept within acceptable bounds. Experiments show promising results for balancing workload. Airspace designs were shown to controllers, and their feedback regarding design considerations were summarized for use in improving our modeling approach.