A Low Latency Clustering Method for Large-Scale Drone Swarms

Abstract

Large-scale drone swarms are expected to play an important role in information acquisition, rescue search and joint fire strikes. These swarms usually adopt a clustering structure to control formation flight for fast and stable communication. Effective clustering can improve the transmission efficiency and task execution ability of the network. On the basis of uniform clustering, we establish a model with the number of unmanned aerial vehicles (UAVs) and the number of cluster heads (CHs) as variables to minimize the communication latency. Using conditional criteria, the communication delay is minimized to solve the relationship between the number of drones and the number of CHs, and this model is for clustering. A simulation platform is built with OPNET to evaluate the impact of different numbers of CHs on the network performance. According to the proposed scheme, the optimal number of CHs for 500, 300, 100 UAVs is 31, 24, 14, respectively. In the case of specific simulation parameters, these optimal numbers of CHs can achieve excellent performance in terms of delay and packet loss rate. This result has value in drone swarms clustering applications.