Self-Organizing Multi-User UAV Swarm Simulation Platform

Abstract

Unmanned aerial vehicles (UAV) swarms offer a cost-effective, time-efficient data collection and analysis solution across various applications. The study presents a cutting-edge self-organizing UAV swarm simulation platform empowered by collective artificial intelligence designed to facilitate terrain monitoring and optimize task performance using a fleet of UAVs. The cloud-based multi-user platform provides users with interactive features for seamless user collaboration and real-time video viewing for collective exploration of dynamic terrain imagery, allowing users to generate requests seamlessly from the QT interface. The UAV map configurator facilitates the creation and modification of UAV swarm navigation maps, optimizing their behavior and performance. Additionally, the parameter gossip system fosters communication and coordination among swarm members, while the QT service layer ensures secure data transfer to cloud servers. This integrated data fuels the formation of essential swarm and target tasks, determining key parameters such as swarm participant count, initial relative coordinate positions, and statuses (imager and/or strike). The server employs advanced algorithms to achieve these functionalities, including the research road graph based on the rotor-router model and the comprehensive information exchange graph using the gossip/broadcast model. These algorithms work synergistically within the server environment, enabling efficient task planning and coordination among the UAV swarm. Furthermore, the platform allows for the seamless transmission of the formed target tasks to the memory of individual swarm participants, enhancing their decision-making capabilities and overall swarm performance.