Multi model transmission analysis based efficient intrusion detection system for improved performance

Abstract

The problem of intrusion detection in network systems has been well studied. There exist numerous techniques in the mitigation of intrusion attacks, but they struggle to produce expected performance. To solve this issue, an efficient multi model analysis based approach is described in this article. The network systems faces various challenges like modification, distributed denial of service, spoofing, eavesdrop and so on. The proposed multi model approach monitors the network packets in different level by analyzing the payload, path, host and frequency of incoming packets. The method considers the frequency of packets, path being used, and frequency of transmission, host details and payload features. For each features, the method computes the trust measure which has been used to classify the packets. The method estimates cumulative multi mode trust weight towards any packet being received. According to the weight measures of different analysis, the attack has been identified. The proposed method improves the performance of intrusion detection and increases the accuracy.