2019 · 17 citations · 7 references
ComputingComputer ScienceComputer SystemsArtificial IntelligenceMachine LearningMobile ComputingDeep LearningFault DetectionMobile CommunicationMobile NetworkingMobile NetworksWireless Networks5G Mobile Networks5G NetworksDeep Learning-based Anomaly DetectionDeep Learning-based DetectionDeep Learning-based MonitoringFault Detection Using Deep LearningMachine Learning DiagnosticsMachine Learning-based Monitoring
This paper introduces a novel AI framework for nationwide wireless KPI monitoring and its diagnosis in LTE/5G mobile networks. The proposed method, which has not been provided so far for a commercial use, automatically detects meaningful degradations of cell-specific wireless KPI statistics and also provides an alarm to administrator with possible causes of the corresponding problem. Machine learning and deep learning technologies are properly applied to detect relative KPI degradations on the basis of those estimated globa/local PDFs and also to carry out root cause analysis for the problem diagnosis, respectively. We have confirmed the feasibility of the mentioned AI framework by completing a first PoC through interworking our AI server with commercial LTE/5G networks. Currently, about 20 major monitoring KPIs such as RRE, ENDC,RSSI, BLER, and so on are being considered in current 5G NSA mode.
7
Automatic differentiation in PyTorch
Adam Paszke, Sam Gross, Soumith Chintala et al. · 2017 · 11.1K citations
Information theory and statistics
Journal of the Franklin Institute · 1959 · 7.2K citations