Optimized Support Vector Machine Based Fused IoT Network Security Management

The emerging elevation of data generation, gathering, and processing has explicitly increased the attacks that become congenital issues in the computer network. The Internet of Things become an essential technology nowadays for processing the above-mentioned process. The security management of IoT become intricate and in our work, we have introduced data fusion technology which comprises data from various sources to establish precise, accurate, and informative information. Meanwhile, the previous works suffered from overfitting and non-optimized intrusion detection. To tackle this, the data fusion is effectuated with the horizontal emergence of two various datasets such as IoT-23 and IoTID20. The fused dataset is forwarded to the proposed support vector machine (SVM) based Binary Grasshopper Optimization algorithm (BGO). The proposed BGO-SVM detects the intrusion in the fused dataset as normal and abnormal. Simulation is effectuated and the performance of the proposed work is analyzed with the existing works in terms of detection rate, accuracy, recall, and F-measure. The proposed work ensures better network security management than the state-of-art works.