Network Attack Identification through Intelligent Model
Detection of the network attack proficiently needs to capture the large amount of the traffic as the dump that needs to be studied. This implies that the very huge amount of the network traffic generated is collected from the transactions that takes place through the network. The identification of the network attack is then performed on this network traffic thus obtained. The identification of the network attack is similar to that of the intrusion into the system which is obtained from the analysis of the data from the traffic of the network. The intelligent approach is thus needed to find the intrusion from the large amount of the dump data which makes the predictions similar to that of the intrusion detection system. For the same the NSL-KDD shall be used for the experimental purpose as it is incorporated with the large amount of the data from network, features, testing dataset, training dataset etc. in this paper the hybrid algorithm is developed which is based on generating less false alarm rate, and it can withstand with the threshold level of intrusion identification from the predefined datasets and this is based on optimized features obtained through the process of preprocessing of the dataset. The hybrid algorithm shall be having the enhanced time complexity, computational speed, efficient identification of the network based attack. The proposed hybrid algorithm shall also be dealing with the issues of the false positive alarm and negative rates. In the proposed hybrid algorithm firstly the data of network traffic is refined using the vote algorithm and then proposed hybrid algorithm is the combination of algorithms like naïve bayes, random tree, and many more.
Authors: Richa Sharma, Priyanka Sharma
Published in: World Congress on Internet Security (WorldCIS-2020)
- Date of Conference: 8-10 December 2020
- DOI: 10.20533/WorldCIS.2020.0002
- ISBN: 978-1-913572-24-2
- Conference Location: London, UK