AN IMPROVED PSO AND RANDOM FOREST FOR INTRUSION DETECTION IN SOFTWARE DEFINED NETWORKS
C.ASWINI, Dr. M. L. Valarmathi, M. Nivedha, T. Ponsheka
Binary Bat Algorithm, Swarm Division, Random Forest, Improved Particle Swarm Optimization
With the exponential growth of network for huge amount of data transmission, there exist an equal chance of network security issues as well. Software Defined Networks takes care of the network architecture intelligently and also controls them with software application. To make it more effective Intrusion Detection System (IDS) goes in hand with improved features to identify the network anomalies precisely with the help of machine learning concepts. This paper uses the Binary Bat algorithm for the selection of features which is done with the help of swarm division mechanism. Further for the process of flow classification, weighted voting mechanism has been used by altering the sample’s weight by the Random forest method. The flow is classified intelligently with selected features to gives better performance result. Evaluation results prove that the modified intelligent algorithms select more important features and achieve superior performance in flow classification. It is also verified that the proposed system shows better accuracy with lower overhead compared with existing solutions. Thus the proposed system helps us in achieving more accurate results when compared to our existing methods.