Software-Defined Networking (SDN) redefines network management by separating
the control and data planes, enabling centralized programmability for emerging applications such as 5G, IoT, and edge computing. Despite its advantages, SDN faces challenges including controller vulnerabilities, limited flow table capacity, and insufficient adaptability to rapidly changing threats. Machine Learning (ML) offers a
promising approach to address these limitations by enabling intelligent, data-driven
decision-making across SDN environments. This thesis conducts a systematic review of 50 peer-reviewed studies from 2020 to 2025, examining how ML is applied
to SDN in the areas of security, traffic engineering, resource management, and Quality of Service (QoS) optimization. A structured taxonomy is presented, organizing ML paradigms—supervised, unsupervised, deep learning, and reinforcement learning—and mapping their applications to specific SDN functionalities. The review reveals that supervised learning techniques are frequently employed for SDN security tasks, while reinforcement learning is effective in routing optimization and deep learning performs well in modeling complex or emerging threats. Key challenges identified across the reviewed literature include a reliance on simulated environments, limited resilience to adversarial inputs, and the computational demands of computationally intensive models, all of which may impede real-world deployment. This thesis outlines research gaps and recommends future directions, including the development of standardized datasets, privacy-preserving ML techniques, and greater use of real-world testbeds to improve practical applicability. By synthesizing the current state of MLintegrated SDN research, this work offers a strategic framework for advancing secure, scalable, and intelligent network systems.
Use Your Cell Phone as a Document Camera in Zoom
From Computer
Log in and start your Zoom session with participants
From Phone
To use your cell phone as a makeshift document camera
