On the Fly Topological Formation of UAV Swarms
Abstract
With over 600 thousand unmanned aerial vehicles (UAVs) set to take flight within the year, the next wave of autonomous defense applications such as machine to machine combat, search and rescue missions, high definition reconnaissance, and rapid connectivity for disaster recovery requires coordination among UAVs to form three dimensional (3D) topologies on the fly. These topologies depend largely on the intermediate and surrounding geographical features and dynamic spectral activity in a given environment. Fortunately, UAVs are often equipped with a large number of sensors such as Inertial Measurement Units (IMUs), Light Imaging, Detection, and Ranging (LiDAR), or sophisticated high resolution (4K) or even 3D cameras that can interpret their surroundings. In this project, we seek to establish a relationship between geographical features observed through images, video, and LiDAR with wireless channel characteristics to establish robust and high performance communication links for on the fly topological formation of UAV swarms. In effect, we propose to build a multi dimensional, multi modal data processing framework that includes signal filtering to improve integrity in noisy conditions, extracting information when outliers exist according to contextual factors, and identifying gaps in data to inform future autonomous UAV reconfiguration. Our results indicate that we can predict and verify a doubling in the estimate of the path loss exponent for locations within a few meters from each other, which would dramatically change an optimal swarm topology. By understanding and quantifying these spatially varying estimates on the fly will allow us to develop efficient swarm topologies under unknown field conditions.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 14, 2022
- Source ID
- FA95501910375
Entities
People
- Joseph Camp
Organizations
- Air Force Office of Scientific Research
- Southern Methodist University
- United States Air Force