{"id":1011078,"date":"2024-03-01T08:39:49","date_gmt":"2024-03-01T16:39:49","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-video&p=1011078"},"modified":"2024-03-15T05:31:30","modified_gmt":"2024-03-15T12:31:30","slug":"project-janus-the-anomaly-detection-demo-mwc-2024","status":"publish","type":"msr-video","link":"https:\/\/www.microsoft.com\/en-us\/research\/video\/project-janus-the-anomaly-detection-demo-mwc-2024\/","title":{"rendered":"Project Janus: The anomaly detection demo (MWC 2024)"},"content":{"rendered":"
Explore the transformative world of vRAN operations in this insightful video, where frequent updates and complex troubleshooting meet the power of AI and machine learning. Discover how defining a normal operational state in vRAN is challenging due to real-time constraints and how generative AI models are revolutionizing problem detection by learning and identifying deviations proactively. We demonstrate a distributed unsupervised machine learning model that collects extensive telemetry and deploys real-time anomaly detection, using over 600 metrics and edge computing for precise issue resolution. Witness how this AI-driven approach simplifies the “single point of management” dilemma, automates troubleshooting, and promises a future of optimized wireless networks with unparalleled efficiency. This is a joint demo with the University of Edinburgh, running on top of Capgemini 5G RAN with Intel FlexRAN.<\/p>\n