Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore an 18-minute conference presentation introducing NetKeeper, an autonomous network configuration update framework designed to enhance network resilience through intelligent traffic pattern analysis and anomaly detection. Learn how this innovative system addresses the limitations of existing configuration methods by incorporating intent understanding, traffic analysis optimization, and dynamic network adaptability to simplify overall configuration management. Discover the framework's multimodal network intent capabilities that process both natural language input and anomaly logs, enabling real-time adaptability to network dynamics. Examine the implementation of northbound and southbound interfaces that effectively translate network intents from operators and network management platforms, creating a seamless bridge between network intentions and actual network behaviors. Understand the multi-agent reinforcement learning model architecture that divides agents based on configuration parameter types to achieve optimal network resilience while maintaining forwarding policy satisfaction. Review experimental results demonstrating NetKeeper's effectiveness in dynamic network environments, including 99.6% average policy consistency, 5.3% improvement in network performance, and 8.7% reduction in traffic shift on average.
