HawkFish optimization-enhanced convolutional neural network for high-accuracy intrusion detection in iot networks

The rapid expansion of the Internet of Things (IoT) has introduced a highly heterogeneous and resource-constrained ecosystem that remains vulnerable to a wide range of cyberattacks. Traditional intrusion detection systems (IDS) struggle to maintain high accuracy while adapting to dynamic IoT environments. To address these challenges, this paper proposes a hybrid intrusion detection framework that integrates the HawkFish Optimization Algorithm (HFOA) with a Convolutional Neural Network (CNN) to enhance detection performance and improve model generalization. The HFOA is employed to optimize critical CNN hyperparameters, including learning rate, convolutional depth, filter size, and regularization strength, ensuring an adaptive exploration–exploitation balance during training. The optimized CNN is then evaluated on benchmark IoT intrusion datasets, demonstrating superior accuracy, detection rate, and robustness compared with conventional machine learning and deep learning baselines. Experimental results show that the HFOA-enhanced CNN significantly reduces false positives while improving classification reliability, making it a strong candidate for real-time IoT network protection. This hybrid method provides a scalable and energy-efficient solution suitable for modern IoT deployments.