Improving Intrusion Detection with Fused IGAN-IDs and Randomized Tree Classification for Enhanced Performance
Kiran Sree Pokkuluri, Sandeep Kumar Awasthy, Vikas Sarkar, Nirav V. Bhatt, Jangam Subbarayudu, A S M Udayakumar · 2025
This paper introduces iGAN-IDS with Randomized Tree Classification, an improved performance method of IDS. The iGAN–IDS uses a GAN to predict anomalies in network traffic; the IDS obtains knowledge that helps it to protect against new threats. In general, due to the exploitation of a set of positive randomized decision trees, the Randomized Tree Classification enhances the probability of the detection of intrusions. What has been proposed is the system that can detect anomalies, generate few false alarms and do it on large scale when it is coupled with integrated approach. Compared with other kinds of IDS methods the suggested IDS approach has better detection rates, more flexibility and less computational complexity as the simulation results on crucial benchmarks datasets prove. From the study, it was revealed that the technology for protection from modern threats within the network systems is more efficient than had been envisaged. The presence of iGAN-IDS and RTC confirms the extension of the existing understanding of the intrusion detection system and can become the subject of further research that will be aimed at making such systems intelligent and autonomous.