5G 시스템에서 Beam RSRP를 활용한 GNN 기반 측위 기법
GNN-based Localization with Beam RSRP in 5G Systems
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
In this study, we propose a Graph Neural Network (GNN)-based localization approach utilizing beam Reference Signal Received Power (RSRP) measurements in 5G networks. Existing GNN-based localization approaches have mainly targeted indoor environments with dense Base Station (BS) deployments and are less effective in outdoor scenarios where the number of BSs is limited. To overcome this limitation, we construct an expanded graph extracted by features such as RSRP, angle, and Line of Sight/None-Line of Sight (LoS/NLoS) indicators, assuming that such information is available at the User Equipment (UE). We validate the proposed approach through ray-tracing simulations in a realistic outdoor 5G environment and evaluate its localization accuracy compared to the triangulation approach.
Abstract
In this study, we propose a Graph Neural Network (GNN)-based localization approach utilizing beam Reference Signal Received Power (RSRP) measurements in 5G networks. Existing GNN-based localization approaches have mainly targeted indoor environments with dense Base Station (BS) deployments and are less effective in outdoor scenarios where the number of BSs is limited. To overcome this limitation, we construct an expanded graph extracted by features such as RSRP, angle, and Line of Sight/None-Line of Sight (LoS/NLoS) indicators, assuming that such information is available at the User Equipment (UE). We validate the proposed approach through ray-tracing simulations in a realistic outdoor 5G environment and evaluate its localization accuracy compared to the triangulation approach.