Detection of Network Intrusion Applying Generative Adversarial Networks

In the world of cyber security, the network traffic monitoring tool, Intrusion Detection System (IDS), as undoubtedly the most important, is continuously active and hunting for security breaches. The number of online attacks creates a market for the smart and automated technology to discern malicious activities. Although, traditional machine learning methods, being precise, are plagued with the issue of imbalanced datasets that make the detection rates to be low and generate a number of false alarms. We therefore plan on cGAN integration which is an innovation from traditional approaches. The cGAN model, using NSL-KDD dataset as an evaluation tool, shows that the model is more superior to the old GAN, developing an appropriate solution for intrusion identification. Our method is centered at improving the reliability of the system, yet also considers the necessity of computational resource consumption. The experiment outcomes prove the usefulness of the cGAN model in fighting with computer assaults which in the end is one of the key factors to provide safety to computer networks.