A game theoretic self-organization for satellite-based optical sensor allocation

Abstract

To aid the development of space situation awareness, this research investigates sensor allocation problem in a multi-satellite system, proposing an approach for self-organized space object tracking. The approach is based on game theory, where the system-level model of the sensor allocation problem is formulated as a potential game, and each satellite-based sensor in the system is treated as a rational player. Each player in the game updates its action by coordination adhering to the synchronous best response rule. This coordination process is proved to converge in a finite number of steps to a Nash equilibrium that is a feasible solution of the system-level model. The performance of the proposed methodology was compared with a benchmark auction-based method through numerical simulation. In the simulated scenario, the proposed method achieved 49.8-65.2% improved tracking precision, and 63.1% improved convergence speed.