On-the-Fly NLoS Detection for Wireless Positioning: Combinatorial Data Augmentation Approach

초록

6G demands wireless positioning to be realized in an on-the-fly manner, enabling a user equipment (UE) to be accurately and instantaneously localized without relying on predetermined statistical models or long-term datasets. The absence of such prior information makes it difficult to mitigate the unpredictable bias caused by non-line-of-sight (NLoS) links. To address this challenge, we propose a novel NLoS detection algorithm, termed combinatorial data augmentation-guided NLoS detection (CDA-ND), which builds upon our prior work. CDA-ND generates numerous preliminary estimated locations (PELs) from different gNodeB (gNB) combinations. When a target gNB is in NLoS, the resulting PELs naturally split into two clusters: one derived using the target gNB's distance measurement, and the other derived without it. The deviation between the two clusters is represented by a single vector, called the NLoS evidence vector (NEV), which serves as a key feature for computing an NLoS likelihood score. The proposed CDA-ND achieves high reliability in indoor factory environments, attaining precisions of 81.14% and 88.28% when the proportion of NLoS gNBs is 18% and 56%, respectively. As a result, a positioning algorithm integrating CDA-ND significantly improves positioning accuracy, achieving a 47.36% reduction in mean absolute error under an NLoS-dominant condition. © 2026 IEEE.

키워드