A Hybrid Algorithm for Modifying and Tracking Connectivity in UAV Teams

Abstract

Algebraic connectivity is the second-smallest eigenvalue of the Laplacian matrix and can be used as a metric for the robustness and efficiency of a network. This connectivity concept applies to teams of multiple unmanned aerial vehicles (UAVs) performing cooperative tasks, such as arriving at a consensus. As a UAV team completes its mission, it often needs to control the network connectivity. The algebraic connectivity can be controlled by altering edge weights through movement of individual UAVs in the team, or by adding and deleting edges. The addition and deletion problem for algebraic connectivity, however, is NP-hard and caused multiple heuristic methods to be developed. A leading method, the greedy perturbation heuristic, is efficient but not always effective. An alternative method, the bisection method, is highly effective but less efficient. The primary contributions of this paper are identification of a set of features and a classifier for predicting when the greedy perturbation heuristic is successful, and presentation of a hybrid algorithm which combines these two methods to provide both effectiveness and efficiency.