DefinePK

Abstract

Network security is one of the primary concerned now days for large organizations. MANETs are deployed in challenging real-world environments for example, disaster management, civilian emergency operation and unmanned aircraft operation. Providing Security in this type of environment is the most challenging problems as the operation environment of such network is usually is difficult to decide. This will lead to any malicious activity that will affect the operation of the networks. Various modifications are being applied to IDS regularly to detect new attacks to handle them effectively. In this paper, we have designed a effective Intrusion Detection System (IDS) to detect the attacks in a MANET that uses modified fuzzy min max neural network in a effective manner. The aim of the proposed work is to improve the detection accuracy ratio and reduce the false positive rate. IDS is designed as a two phases technique, in which the first phase use the fuzzy min max neural network algorithm to train 60% data and test 40% data of the network. In the second phase, the modified fuzzy min max neural network algorithm is applied to 60% of train data +20% test data to form a new train data set and test the remaining 20% data collected. The proposed two phase intrusion detection method was evaluated using KDD cup data set. The experimental results showed that the proposed method is better than the other methods in terms of the detection rate, accuracy and reduced false positive rate.