A Novel SIP-Based Traffic Feature Extraction Model Using an Adaptive Approach in VoIP Environment

Due to its low usage and deployment costs, flexible communication, high scalability, and ease of integration, VoIP (Voice over Internet Protocol) has been widely applied in various Internet communication scenarios, such as homes, enterprises, and service providers. It provides the technological foundation for modern, high-quality, interconnected, and multifunctional communication services. In this context, the security and stability of VoIP systems are of paramount importance for the development of current society. The SIP (Session Initiation Protocol) serves as the foundation for managing most of the VoIP communication data and it faces various novel network attacks. Therefore, this paper proposes an adaptive flow feature extraction method that filters, extracts, and analyzes key features from SIP messages. This approach narrows down the scope of feature research, facilitating more effective network data analysis. By optimizing the user communication experience, identifying abnormal activities in the network, and enhancing the security and reliability of VoIP networks, this method contributes to a safer and more stable environment.