CMAC-based API Security with Transformer Model for Intrusion Detection in SDN-IoT Networks

This paper proposes a secure API framework for SDN-IoT networks by utilizing a Transformer-based intrusion detection model integrated with CMAC for robust API security. The Yule-Simon Distribution-Based Lyrebird Optimization Algorithm (YSD-LOA) is leveraged to optimize feature selection and reduce both encryption and detection times while enhancing accuracy and security. This framework merges the Transformer model for effective anomaly detection with CMAC to ensure the secure communication of SDN controllers and IoT devices. The proposed system achieves remarkable intrusion detection accuracy of 99.2% and outperforms conventional methods such as HMAC and RNN by reducing delay by 90ms and encryption time by 1050ms on average. This optimized advanced combination of CMAC and Transformer by YSD-LOA, being scalable, efficient, and secure for intrusion detection in real-time, can be perfectly fitted for the next generation of SDN-IoT networks both in scalability and security.