Intrusion Detection System Based On Machine Learning Algorithms In the DOUNG Ad Hoc Network

The New Generation of Open Digital Universities (DOUNG) [1] is a model designed to cover as few human resources (professors) and material resources (lecture halls and classrooms) as possible. Its architecture offers global ubiquity and tailor-made offers to learners by relying on the IT and telecom networks already available. Thus, it inherits the vulnerabilities and threats inherent in these channels, in particular those of wireless networks and without pre-existing infrastructure. This lengthens the perimeter to be protected and exponentially increases the risk of intrusions into the synchronous course offer. To overcome these challenges, we proposed and tested an intelligent intrusion detection model based on machine learning (ML). Based on the approaches used (Fl score and confusion matrix), it appears from our experiment that logistic regression (LR) provides the best detection of MiTM attacks [1]–[3]. in the ad hoc network of the DOUNG.