A hybrid machine-learning based security algorithm for improved detection of distributed-denial-of-services attacks in internet of things networks

The Internet of Things (IoT) has emerged as a critical enabler of intelligent and autonomous systems, supporting applications across sectors such as agriculture, healthcare, education, and smart infrastructure. However, the increasing connectivity of IoT devices has significantly expanded the attack surface, making these networks particularly vulnerable to Distributed Denial of Service (DDoS) attacks. Existing detection approaches, including Decision Tree (DT), Linear Regression (LR), and K-means, are often limited by suboptimal classification accuracy, high response latency, and poor adaptability to evolving and zero-day attack patterns, thereby compromising their effectiveness in dynamic IoT environments. To address these limitations, this study proposes a Machine Learning-Based Security (MLBS) framework that integrates K-Nearest Neighbors (KNN) and Recurrent Neural Networks (RNN) to enhance the detection and mitigation of DDoS attacks in IoT networks. The proposed hybrid model leverages the temporal learning capability of RNNs and the instance-based classification strength of KNN to improve both detection accuracy and generalization. The framework was evaluated using the MATLAB simulation environment, where it achieved an accuracy of 98.72%, precision of 97.95%, recall of 98.94%, and an F1-score of 98.39%, outperforming conventional baseline models. Beyond performance improvements, the proposed MLBS framework demonstrates strong performance on previously unseen traffic patterns, indicating its suitability for real-time deployment. The key contribution of this work lies in providing a scalable, adaptive, and high-precision intrusion detection solution that enhances the resilience of IoT networks against both known and emerging DDoS threats.