Scattering Modeling of Swarm Targets Using Scattering Center Representation and Physics-Informed Neural Networks

Accurate modeling of electromagnetic scattering from swarm targets is crucial for developing effective countermeasures against coordinated threats. This paper presents a scattering center (SC) modeling framework based on the geometric decomposition of target components, enabling efficient electromagnetic characterization of swarm targets. The proposed method analytically represents the structural features of individual targets and incorporates a ray-based hybrid approach for occlusion management, significantly reducing both the number of SCs and the computational cost. To account for inter-target coupling effects, a physics-informed neural network (PINN) is integrated into the SC model to rapidly predict representative coupling structures. Numerical results validate the accuracy of the proposed approach across various swarm configurations, achieving performance comparable to full-wave methods while substantially reducing computational time. These capabilities make the framework well-suited for fast electromagnetic simulation of swarm-based scenarios.