Network Anomaly Detection Using Artificial Neural Networks Optimised with PSO-DE Hybrid

• Published in: Security in Computing and Communications Vol. 969; pp. 257 - 270
• Main Authors: Rithesh, K., Gautham, Adwaith V., Chandra Sekaran, K.
• Format: Book Chapter
• Language: English
• Published: Singapore Springer Singapore Pte. Limited 2019; Springer Singapore
• Series: Communications in Computer and Information Science
• Subjects: Anomaly-based NIDS; Differential Evolution; Network traffic; Neural network; Stream data analysis; Swarm optimiser
• ISBN: 9811358257; 9789811358258
• ISSN: 1865-0929; 1865-0937
• DOI: 10.1007/978-981-13-5826-5_19

Anomaly Detection is an important field of research in the present age of ubiquitous computing. Increased importance in Network Monitoring and Security due to the growing Internet is the driving force for coming up with new techniques for detecting anomalies in network behaviour. In this paper, Artificial Neural Network (ANN) model optimised with a hybrid of Particle Swarm Optimiser (PSO) and Differential Evolution (DE) is proposed to monitor the behaviour of the network and detect any anomaly in it. We have considered two subsets of 2000 and 10000 dataset size of the NSL KDD dataset for training and testing our model and the results from this model is compared with the traditional ANN-PSO algorithm, and one of the existing variants of PSO-DE algorithm. The performance measures used for the analysis of results are the training time, precision, recall and f1-score.