An Efficient IoT Security Prediction Framework

The Internet of Things (IoT) is widely used in modern industrial applications due to its inherent capability of transferring data without human intervention. A Wireless Sensor Network (WSN) is the central part and plays an important role in IoT. These are widely used in complex applications in which various security attacks are encountered to prevent and countermeasure the problems in WSN, such as data processing, analysis, and efficient utilization of network resources. To ensure secure data processing, techniques are proposed to protect data privacy. Therefore, we propose a novel data-driven model using the benchmark NSL-KDD dataset based on a randomized tree, which improves the performance of WSNs. The performance analysis of the proposed model has been analyzed by using different metrics, e.g., RMSE, MSE, MAE, R-Squared, training time, model size, and prediction speed. The results obtained in the experiments depict that the proposed model effectively performs better than other contemporary models in the state of the art in case of different errors such as RMSE, MSE, MAE, R-Squared, etc.