QoS-Aware scheduling policies for Open Fronthaul transport networks

Abstract

We have recently seen a paradigm shift on the architecture for Radio Access Networks (RAN), being Open-RAN (O-RAN) architecture one of the most relevant examples for 5G and 6G networks. In this work, we combine theoretical models, based on the 5G New Radio standard, and cross-validation through real experimentation, using srsRAN software, to characterize Open Fronthaul traffic. Then, a realistic simulation framework based on ns-3 is designed and implemented to assess the performance of various scheduling policies for intermediate nodes providing fronthaul connectivity, focusing on the integration of Quality-of-Service (QoS) strategies for serving fronthaul together with other services (in the form of 5G slices) on those nodes. By exploiting such methodology, we assess the benefits of prioritization schemes, highlighting the need of Hierarchical QoS solutions, particularly for congestion events. We carry out an extensive simulation campaign, and we show that using appropriate scheduling solutions yields relevant benefits to slices having tight delay requirements, compared to baseline solutions, reducing not only the average value, but also its variability.