DG-IRFC: Intelligent design and implementation of network intrusion detection system (NIDS)

In a rapidly evolving digital environment, ensuring secure communication remains a growing challenge as threats become increasingly pervasive. Intrusion Detection Systems (IDS) help maintain net integrity by identifying and sensing unauthorized or malicious activities. Conventional IDS architectures frequently produce erroneous findings and are unable to modify their definitions of attacks to reflect the changing nature of assaults. The goal of this research is to use the Dynamic Gravitational-tuned Intelligent Random Forest Classifier (DG-IRFC) to develop and deploy a revolutionary Machine Learning (ML)-based Network Intrusion Detection System (NIDS) to overcome these issues. The proposed framework utilizes a combination of public intrusion detection datasets and real-time network traffic collected from simulated environments. Preprocessing included handling missing values and applying Z-score normalization to balance and refine input data, enhancing classification accuracy and stability in the intrusion detection process. Kernel-Principal Component Analysis (Kernel-PCA) is used for dimensionality reduction and relevant feature withdrawal. The DG-IRFC algorithm integrates gravitational search principles with intelligent parameter tuning within the Intelligent Random Forest structure, enabling dynamic adaptation to complex traffic behaviors. The classifier enhances multi-class intrusion detection efficiency by minimizing overfitting and improving decision boundaries. The proposed NIDS architecture processes traffic in real time, extracts relevant features, and classifies traffic into attack categories with improved precision. Evaluation results confirm the superiority of DG-IRFC, with 98.85 % accuracy, 98.85 % recall, 98.50 % precision, 98.47 % F1-score, and 72.84 s Inference time, reflecting its efficiency in reliably recognizing and classifying network intrusions across varied attack patterns. This approach offers a clever and scalable description of the structure of modern web security.