LossLeaP: Learning to Predict for Intent-Based Networking

Intent-Based Networking mandates that high-level human-understandable intents are automatically interpreted and
implemented by network management entities. As a key part in this process, it is required that network orchestrators activate
the correct automated decision model to meet the intent objective. In anticipatory networking tasks, this requirement
maps to identifying and deploying a tailored prediction model that can produce a forecast aligned with the specific –and
typically complex– network management goal expressed by the original intent. Current forecasting models for network demands
or network management optimize generic, non-flexible, and manually designed objectives, hence do not fulfil the needs
of anticipatory Intent-Based Networking. To close this gap, we propose LossLeaP, a novel forecasting model that can
autonomously learn the relationship between the prediction  and the target management objective, steering the former to minimize
the latter. To this end, LossLeaP adopts an original deep learning architecture that advances current efforts in automated
machine learning, towards a spontaneous design of loss functions for regression tasks. Extensive experiments in controlled
environments and in practical application case studies prove that LossLeaP outperforms a wide range of benchmarks, including
state-of-the-art solutions for network capacity forecasting.