arXiv:2602.20018v2 Announce Type: replace Abstract: Softwarized radio access networks (RANs), such as those based on the Open RAN (O-RAN) architecture, generate rich streams of key performance indicators (KPIs) that can be leveraged to extract actionable intelligence for network optimization. However, bridging the gap between low-level KPI measurements and high-level requirements, such as quality of experience (QoE), requires methods that are both relevant, capturing temporal patterns predictive of user-level outcomes, and interpretable, providing human-readable insights that operators can validate and act upon. This paper introduces conformal signal temporal logic learning (C-STLL), a framework that addresses both requirements. C-STLL leverages signal temporal logic (STL), a formal language for specifying temporal properties of time series, to learn from data interpretable formulas that characterize key properties of KPI traces satisfying high-level requirements. We first formulate the problem of reliably identifying an accurate STL formula as a risk-averse inference problem. We show that the optimal solution requires the construction of a set of STL formulas with formal coverage guarantees. Based on this theoretical insight, the proposed C-STLL extracts from calibration data a set of STL formulas with such guarantees via Learn Then Test, while also filtering for formula complexity and diversity. Experiments on a mobile gaming scenario in ns-3 and on KPI traces measured on a softwarized RAN testbed demonstrate that C-STLL effectively controls risk below target levels, while returning interpretable temporal specifications that relate KPI behavior to QoE outcomes.