02 Nov 2019: Dr. Majid Butt gave a seminar about “UE Positioning in a Dense 5G Millimeter Wave Deployment using Supervised Deep Learning”. He discussed about some of the key performance metrics of 5G, such as, ultra reliable low latency communication (URLLC) with reliability requirements of >99% and latency requirements of a few milliseconds. He compared conventional positioning indoor positioning schemes with Machine Learning-based positioning while explaining how well can machine learning in a 5G base station (gNB) predict UE positions using available 5G NR CSI report. He then concluded that a combination of serving and neighboring cell beam reports should be used with at least 2 beams coming from the serving cell and at least one from the neighboring cell. The initial results indicate that this approach can give a localization accuracy of a few meters in the indoor environments. The talk was followed by an interactive discussion and QA session.