期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2018
卷号：7
期号：1
页码：552
DOI：10.15680/IJIRSET.2018.0701039
出版社：S&S Publications
摘要：In this technological age, a lot of data that include secrets have been given away to the internet. Theexponential growth in the E-Commerce, the challenge of security of the secret data is a huge problem to providesecurity to data for individuals, companies and governments. It is very important that the data should reach safely to thedestination and only the intended person should able to access the data without any intrusions over the network. But inthese days, where there is a continuous increase in immorality and decrease in ethical values, cyber crimes havebecome regular news items in our televisions and news papers. Researchers have been contributing various solutionstowards addressing this problem and Network Intrusion Detection Systems (NIDS) is one of the solutions against thesecyber crimes and network attacks. However, the performance of these Network Intrusion Detection Systems needs tobe measured by effective metrics. There are various primitive metrics available such as Accuracy, Detection Rate andFalse Alarms Rate etc., but there is no bench mark performance metric available in the literature to evaluate theperformance of the NIDSs. On the other hand performance metrics play a very major role in data security because theperformance of the NID systems is evaluated through these metrics. In this paper, a novel Composite PerformanceMetric, called ‘DAFAR’ has been proposed to evaluate the performance of different Network Intrusion Detectionsystems. Various experiments were carried out on the benchmark data set (ISCX 2012) [1] to test the efficacy of theNID systems’ performance. The experimental results show that the proposed metric gives better direction whileevaluating the performance of different Network Intrusion Detection systems.
