ABSTRACT
The proposed model integrates GRU-Dropout for real-time IoT-SDN intrusion detection and KMAC for API security. GRU-Dropout prevents over-fitting while KMAC ensures correctness of the data. Ant Colony Optimisation (ACO) algorithm selects relevant features in order to select relevant attributes as it makes it more efficient. To provide KMAC based GRU-Dropout scalable amiable framework for secure API security and high-performance intrusion detection in IoT-SDN networks. It utilizes KMAC to secure the APIs, GRU-Dropout to detect the temporal attack patterns, and ACO to optimize feature selection to improve the detection rate and reduce the computing overhead. 99.49% accurate prediction results within 1200ms of predicted times, outperforming conventional models. The proposed architecture is reliable and real-time in nature and is deemed for protection of large scale IoT-SDN networks with high accuracy and low latency.
