Cognitive networking for uav swarms

Abstract

A new generation of lightweight and small Unmanned Aerial Vehicles (UAVs) enables the design of networked aerial robotic systems for a wide range of applications. In this book chapter challenges and solution approaches with respect to self-organized and robust multi-UAV systems are discussed. In order to achieve a flexible deployment of Unmanned Aerial Systems (UAS), a novel service-oriented system architecture is introduced. The design aspects of mobility control algorithms addressing the competing requirements of communication reliability and spatial distribution of UAV swarms are demonstrated using two reference scenarios with diverging requirements: aerial sensor networks and ad-hoc aerial relay networks. The novel solution approaches presented in this chapter rely on agent-based control of UAV swarms operating autonomously in dynamically changing environments. Different communication aware algorithms for microscopic as well as macroscopic mobility control are presented, such as Cluster Breathing, Smart Cube, Potential Fields, and Role-Based Connectivity Management. For the positioning of UAV relays, the Interference-Aware Positioning of Aerial Relays (IPAR) algorithm is introduced. The system design and optimization of cognitive networking for UAS requires a dedicated multi-scale simulation environment, which includes a detailed physical channel model and real-world experiments. The chapter discusses how the network and application task-specific performance requirements are met with the proposed mobility control algorithms even in the cases of temporarily unavailable communication links. The self-healing capabilities therefore allow for reliable networking and control of aerial robot swarms in diverse use cases, such as emergency response, environmental monitoring, and ad-hoc network provisioning.