Cyber-physical systems in critical infrastructure face various threats, prompting research into multiple attack detection mechanisms. Anomaly detectors based on neural networks are one such mechanism. However, they are vulnerable to adversarial attacks, which can compromise their effectiveness. To address this, researchers have proposed various mitigation strategies, including regularization techniques and adversarial training. These strategies aim to improve the robustness of anomaly detectors against adversarial attacks, ensuring their reliability in real-world applications.