Task allocation without direct communication: Graphical game based swarm interception allocation

The interception of hostile UAV swarms in communication-denied environments poses a significant challenge for modern defense systems. Traditional methods, such as physical capture or frequency interference, are ineffective against coordinated swarms, while existing multi-robot task allocation strategies rely heavily on communication. This paper addresses this gap by proposing a game-theoretic framework for swarm interception allocation based on active observation. We formulate the problem as a graphical game, enabling interceptors to make decentralized decisions without direct communication. Specifically, we investigate Correlated Equilibrium (CE) to resolve conflicts in target selection among interceptors. The proposed method achieves task coverage and global payoff comparable to centralized approaches like the Hungarian algorithm, as validated by software-in-the-loop simulations and real-flight experiments. This work demonstrates a robust, communication-free solution for swarm interception in complex scenarios. Video of the experiments is available at youtu.be/twUSc8zJ9so .