Research Article

FNet: A Two-Stream Model for Detecting Adversarial Attacks against 5G-Based Deep Learning Services

Table 5

Parameter setting of three attack methods in adversarial robustness toolbox.

Attack methodParameter

FGSMNorm = 2, eps = 2.0, eps_step = 0.1
DeepFoolEps = 0.1
CWLr = 0.2, confidence = 0.1