Fuzzy based intrusion detection system in MANET

The rapid development and popularization of the Mobile Adhoc Network (MANET) have brought many security issues in network. Intrusion detection system, an effective security technology which can efficiently detect malicious data in complex network environments and ensure computer network security. Because of the complexity of MANET, traditional Intrusion Detection system IDSs are ineffective in this new context, several methods including Support Vector Machine (SVM) have been used to detect intrusion. Most existing technologies strive for low execution times and energy efficiency while achieving accurate detection rates. To overcome these disadvantages, a novel Principal Component Analysis based Fuzzy Extreme learning machine (PCA-FELM) has been proposed in this paper. Initially, the features are extracted by using Principal Component Analysis and then the extracted features are classified by using Fuzzy Extreme Learning Machine. The proposed PCA-FELM is implemented using MAT LAB simulator. The proposed PCA-FELM is compared with existing methods such as DBN-IDS, GOA-SVM and SDAE-ELM and the proposed method achieves higher accuracy of 99.08% than other existing methods. Experiments on the Knowledge Discovery and Data Mining Tools Competition, KDD Cup99 dataset show that the proposed PCA-FELM model have superior performance than other existing techniques. © 2022