CONGEST: An Algorithm to Detect Congestion Zones Based on Unmanned Aerial Vehicle (UAV) Flight plans

One of the primary problems in the Unmanned Aerial System (UAS) field is the potential increase in conflicts and congestion as the UAS gets incorporated into traditional manned airspace. Our research work aimed to come up with an innovative idea to identify potential congestion zones based on the UAV flight plans. A flight plan is a planned route for a flight that contains a set waypoint defined by latitudes, longitudes, and altitudes. Congestion is defined as a circle of radius ‘R’ that has more than a given number of crossings. Conventional clustering algorithms such as DBSCAN, K-means, K-medoids, CLIQUE, etc are designed for static points/objects i.e., the features of the objects remain constant. However, in the case of Unmanned Aerial Vehicles (UAVs), the UAVs keep on changing their position from time to time. Hence, conventional clustering algorithms cannot be applied to the given problem. We have proposed a heuristic algorithm named CONGEST, based on the DBSCAN Clustering algorithm. Our algorithm can identify 91.5-98.5% of the congestion zones in a span of a few seconds.