Imaging technologies are pivotal in the emerging aerial navigation ecosystem. However, these technologies are vulnerable to adversarial attacks. Current methods for enhancing the adversarial robustness of learned models are primarily based on adversarial training. However, such methods show limited success when applied to the high-resolution optical sensors used in aerial navigation. As the robustness of adversarially trained models is known to improve under attacks that closely resemble those seen during training, we suggest enhancing the models’ robustness by focusing on highly realistic attacks. In this work, we aim to evaluate the susceptibility of navigation systems to such attacks and to enhance their corresponding robustness. We first intend to develop a realistic threat model and demonstrate the susceptibility of navigation systems to these attacks. We then aim to present a robust aerial navigation system by enhancing the systems’ robustness via adversarial training incorporating the threat model.