Cooperative Relative Localization for AAV Swarm Based on Multistage Self-Organizing Clustering Algorithm Under GPS Attack

Accurate assisted localization is crucial for autonomous aerial vehicles (AAVs) in the presence of global positioning system (GPS) attacks, as some compromised AAVs lack accurate position information. Existing relative localization methods face challenges, including the redundancy of range and localization errors, which are caused by the large-scale and long-range characteristics of AAV swarms. This article proposes a AAV group cooperative localization strategy based on multistage self-organizing clustering and super multidimensional scaling (SMDS), referred to as M-SMDS. First, to reduce the information range redundancy and improve the communication ability between clusters, the clustering problem is decomposed into a multistage competition between clusters using the idea of the empire game. Given the clustering objective and principle, it has been demonstrated that the proposed method has a stable clustering ability. Second, the M-SMDS combines interclustering and intraclustering range information. Specifically, the distance between each AAV and beacon AAV is inferred based on the relative distance among cluster members, heads, and the cluster group. Simulation results demonstrate that the proposed strategy effectively enhances the communication transmission rate and localization accuracy of the AAV on a large scale, even under various GPS attacks.