This research is developing a framework for better managing IP networks using an intent-based network management approach to address the performance and robustness issues associated with managing complex IP networks. The goal is to automatically generate needed network configurations by requiring only high-level objectives as input while also ensuring network-level objectives such as performance and reliability. The research develops design principles applicable to managing both current and future networks and provides insight into designing networks for manageability. Using abstraction-based representations, the framework enables high-level specifications to automate low-level tasks, so that one can directly manage the network. Three main obstacles in today's network management are addressed: insufficient visibility, inability to predict the outcome of network configuration changes, and hidden errors in configurations due to assembly-language-like interfaces. The research develops these key techniques: (1) Measurement-based derivation of network protocols' operational models to elicit undocumented limiting behavior. (2) Methods to quantify limitations of measurement methodologies to enable more accurate result interpretation. (3) Privacy-preserving, incentive-compatible data sharing across networks. (4) Systematic evaluation of trade-offs by developing metrics for quantifying network properties and abstractions encapsulating device details. (5) Real-time decision support for what-if analysis and automated intent-based configuration generation. (6) Joint control and data plane management.
The research will advance the state of the art in managing IP networks by addressing key challenges in achieving automated, evolvable, and robust network management. Any developed software will be publicly available. The research results will be integrated into undergraduate and graduate curriculum.
