A Review on the Performance Analysis of EAACK, TSTMC, AB-UBTM and HSCT for Intrusion Detection in Mobile Ad-Hoc Network

Abstract

An efficient design of network architecture for intrusion detection using Tuning Spanning Tree Multiclass Classifier (TSTMC) mechanism in MANET is provided. The construction of spanning tree using Tuning Tutte polynomial operation in mobile ad-hoc network improves the classification rate of abnormal nodes in MANET. The spanning tree based classification using breadth first search reduces the time complexity in detecting abnormal activities and significantly reduces the packet delay rate. To enhance the security level for different mobile nodes in ad hoc network, the Uninterrupted Bayesian Time Mobile algorithm is designed. Adaptive Boosting algorithm improves the resource utilization factor using probability of success and failure factor. Next, an Uninterrupted Bayesian Time Mobile (UBTM) is designed depending on the node class state to increase the trust accuracy rate. Hybrid Symmetric Cryptography Technique is provided based on the novel mixture of two symmetric cryptographic techniques using SP-AES algorithm and MD5-MAR for MANET. This technique reduces packet delay time and improves the true positive rate on abnormal activities in Mobile Ad Hoc Network.The experiments of intrusion detection system are conducted for Tuning Spanning Tree Multiclass Classifier (TSTMC) mechanism, Adaptive Boosting with Uninterrupted Bayesian Time Mobile (AB-UBTM) Networks, Hybrid Symmetric Cryptography Technique (HSCT) with the existing method named as Enhanced Adaptive ACKnowledgement (EAACK) illustrated by Elhadi M. Shakshuki., et al., 2013. Some of the parameter used for analyzing the intrusion in mobile networks was security, packet delay time, true positive rate, trust accuracy, resource utilization factor, classification sensitivity rate and multiclass error