The first phase of this internship is to develop knowledge around anomaly detection models, the different categories of attacks for the selected protocol, the use of transformers (BERT, SecBERT, RoBERTa, GPT, etc.) and Generative Adversarial Network (GAN). During the second phase of this internship, the candidate implements the different components of the framework ranging from the design of adversary models, attack generation, traffic analysis, feature extraction and detection. In the third phase, the candidate will develop the software components (protocol) necessary to show the use case, and perform the performance evaluation. In the fourth phase, the candidate will write their internship report.