UAV swarm tactics: an agent-based simulation and Markov process analysis

Abstract

The rapid increase in the use of unmanned aerial vehicles (UAVs) in recent decades lead to their potential use as saturation or swarm threats to Allied Forces. One possible counter measure is the design and deployment of a defensive UAV swarm. This thesis identifies a future concept of swarm-versus-swarm UAV combat, focusing on the implications of swarm tactics and identifies important factors for such engagements. This work provides initial key insights through significant modeling, simulation, and analysis. The contributions of the presented work include the design of an agent-based simulation and the formulation of an associated analytical model. The agent-based simulation allows for the UAV to be modeled as an agent that follows a simple rule set, which is responsible for the emergent swarm behavior relevant to defining swarm tactics. A two-level Markov process is developed to model the air-to-air engagements, where the first level focuses on one-on-one combat while the second level incorporates the results from the first and explores multi-UAV engagements. Tactical insights obtained from this study can be contrasted with tactics for manned air combat, which highlights the potential need to develop new tactics for unmanned combat aviation as well as for swarm scenarios. Additional analysis performed in this thesis provides further tactical recommendations and outlines multiple avenues of future study.