Enhancing Cloud-Native System Robustness: Time-Series GAN-Powered Fault Injection Model

In the realm of cloud-native infrastructures, ascertaining system robustness is crucial for uninterrupted functionality. Cloud-native systems’ dynamic and complex nature often hinders accurate resilience assessments in varied fault scenarios. This research presents an advanced method that leverages Time-Series Generative Adversarial Networks (Time-Series GANs) infused with chaos theory to enhance system resilience through an innovative fault injection model. The experimental results validate our approach, revealing a significant 19.4% improvement in injection accuracy and a 28.6% reduction in training time compared to existing benchmarks.