Towards Integrating Formal Methods into ML-Based Systems for Networking

Owing to its adaptability and scalability, Machine Learning (ML) has gained significant momentum in the networking community. Yet, ML models can still produce outputs that contradict knowledge, i.e., established networking rules and principles. On the other hand, Formal Methods (FM) use rigorous mathematical reasoning based on knowledge, but suffer from the lack of scalability. To capitalize on the complementary strengths of both approaches, we advocate for the integration of knowledge-based FM into ML-based systems for networking problems. Through a case study, we demonstrate the benefits and limitations of using ML models or FM alone. We find that incorporating FM in the training and inference of an ML model yields not only more reliable results but also better performance in various downstream tasks. We hope that our paper inspires a tighter integration of FM-based and ML-based approaches in networking, facilitating the development of more robust and dependable systems.