Ιn the 3GPP RAN3 125-bis meeting in the beautifull Hefei the main discussion is related to:
– RAN slicing on Qos measurements and the exact granularity it should be achieved
– RAN Coverage and Capacity (CCO) on the gNB CU and gNB DU (split architectures) as well as the timing related information for CCO correction actions.
The open question in that meeting is whether the UE performance after CCO should be collected as the feedback information to evaluate the CCO related output or not. There has been proposals like taking the cell throughput, cell level packet loss and cell level packet delay as the feedback information for predicted CCO issue and future CCO state. Another proposal supports to report UE measurement reports as feedback information for predicted CCO issue detection.
There are definately advantages of using UE Performance Feedback for Evaluating CCO Output. For example collecting UE performance metrics can provide detailed insights into the actual experience of end-users post-CCO. The UE performance feedback helps determine whether the CCO changes led to tangible improvements in network conditions from the end-user perspective, offering a more accurate measure of the success of CCO strategies.
Further more the inclusion of UE feedback allows CCO algorithms to adapt dynamically to changing network conditions and user mobility patterns. Apart from that UE measurements can reveal performance variations at the cell edge or in challenging coverage areas. This can help in identifying scenarios where CCO actions might not have been effective, allowing for targeted optimization.
Finally incorporating UE performance metrics into the dataset used for AI/ML-based CCO models can improve the accuracy of future predictions. If a CCO action does not yield the expected improvements, UE performance data can assist in performing root cause analysis, revealing if the issue lies in the CCO algorithm, hardware limitations, or other external factors.
There might be as well asome Different UEs may have varying capabilities when it comes to measurement accuracy and reporting intervals. This inconsistency can affect the quality of the collected data and may skew CCO evaluations, Increased Signaling Overhead, Complexity in Data Correlation and UE performance feedback may take time to be collected and analyzed, potentially leading to delays in evaluating the CCO’s effectiveness and implementing corrective actions.