Federated Learning Integrated CNN-Trans DDoS Attack Detection Model

In DDoS attack monitoring [1] [2], effective detection algorithms [3] that rely on a large amount of data support are needed to achieve the desired detection results. Because individual organisations generate limited amount and characteristics of data, and considering data privacy riskiness they are reluctant to share training data [4], which often results in "data silos", and insufficient data quantity and quality can lead to inaccurate detection results. To address the above issues, this paper will explore how to use joint learning techniques to train models on massive network data while ensuring the privacy and security of organisational data.