CAMTE: a camouflage-aware multidimensional trust evaluation model for LEO satellite networks

To address the threat of advanced internal camouflage attacks in low Earth orbit (LEO) satellite networks, this paper proposes a camouflage-aware multidimensional trust evaluation model. This model abandons the traditional paradigm reliant on indirect trust, instead constructing a direct trust evaluation framework based entirely on objective evidence from three orthogonal dimensions: identity, behavior, and service. This approach inherently immunizes the system against bad-mouthing and collusion attacks. The core innovation lies in proposing a camouflage-aware trust fusion method, which overcomes the limitations of Dempster-Shafer evidence theory by redefining evidence conflict-traditionally treated as noise-as a key signal for identifying node camouflage behavior, thereby enabling precise quantification and detection of advanced camouflage attacks. Simulation results demonstrate that, compared to various existing trust evaluation methods, the proposed model exhibits significant advantages in detection accuracy and threat suppression capability when confronting various camouflage and dynamic attacks, providing a more robust trust evaluation solution for LEO satellite network security.