Latency analysis

A first network entity, e.g. a Network Data Analytics Function (NWDAF), receives from a second network entity a request or subscription for analytics relating to transmission delay associated with one or more UEs; obtains input data from a third network entity, the input data including information relating to the transmission delay; and outputs information relating to the NWDAF-based analytics based on the obtained input data, wherein the information comprises at least one output for the NWDAF-based analytics, and wherein the transmission delay corresponds to traffic to/from the one or more UEs or corresponds to a data volume sent by/to one of the one or more UEs. The transmission delay may relate to traffic associated with AI/ML-based services including AI/ML modelling or training, AI/ML inference, or transfer of AI/ML models, or extended reality and media (XRM). The at least one output may assist with a federated learning operation or another AI/ML operation. The transmission delay-related information may be latency, packet delay distribution, jitter, periodicity, or end-to-end volume transfer time.