Sixth-generation (6G) networks are moving toward AI-native operation, where learning modules are embedded across the radio access network (RAN), edge, and core. This transition requires learning from limited labels, heterogeneous wireless and network data, partial observations, n…
This paper addresses bistatic snapshot radio SLAM, in which a user equipment (UE) with unknown 6-D pose and clock bias is localized and environmental landmarks are reconstructed from a single multipath channel snapshot. Under mixed line-of-sight (LoS)/non-line-of-sight (NLoS) pro…