A Comparative Study of Network Traffic Classification Algorithms in SDN

Software-defined networking (SDN) has transformed network architecture by separating the control plane from the data plane, allowing centralized control and dynamic programmability. This architectural shift has made network traffic classification a crucial task for optimizing network performance, enforcing quality of service, and enhancing security. Traditional classification methods, such as port-based and deep packet inspection, face challenges in handling encrypted and large-scale traffic. The emergence of machine learning (ML) and deep learning (DL) algorithms has significantly improved traffic classification in SDN by utilizing flow-based features and adaptive learning techniques. This study presents a comparative analysis of various ML and DL-based traffic classification algorithms, evaluating their efficiency, scalability, and applicability in SDN environments. Among these, DL-based algorithms demonstrate superior performance in handling complex datasets. The analysis highlights the effectiveness of ML and DL algorithms in improving classification accuracy and real-time network adaptability.