Coded Distributed Path Planning for Unmanned Aerial Vehicles
Cameron Douma, Baoqian Wang, Junfei Xie · AIAA AVIATION 2021 FORUM · 2021
View Video Presentation: https://doi.org/10.2514/6.2021-2378.vid To enable urban air mobility (UAM), an efficient shortest path planning algorithm is required to ensure safe UAV navigation in large-scale urban environments. Existing optimal shortest path planning algorithms, e.g., the Dijkstra's algorithm, are computationally infeasible for complicated scenarios and large-scale problems. This paper aims to conquer this challenge by exploring a novel distributed implementation of the classical centralized Dijkstra's algorithm. The proposed algorithm explores the coding theory and the idea of load balancing to address the practical issues prominent in UAV-based distributed computing systems, including uncertain disturbances and node heterogeneity. Comprehensive experimental studies on Amazon EC2 demonstrate the high resiliency of the proposed algorithm to uncertain disturbances and its high efficiency compared with existing solutions.