http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Near-Field Localization for Hybrid Beamforming Systems
Suhwan Jang(장수환),Chungyong Lee(이충용) 대한전자공학회 2024 대한전자공학회 학술대회 Vol.2024 No.6
By This paper presents localization method for the near-field user equipment (UE) in hybrid massive multiple-input multiple-output (MIMO) systems. In general, the far-field assumption, where wavefronts make angle-dependent planar wave, is made in the system model. However, the effect of near-field becomes dominant in massive MIMO systems. In the near-field region, the wavefronts form the location-dependent spherical wave. Hence, we the values of received pilot dependent on the range, which means that the localization can be achieved using the received signal. To localize the UE using the maximum-likelihood (ML)-based objective function, we pre-white the received pilot and alters complex channel gain into location-dependent value. Then, we exhaustively search locations domain for the localization using the corresponding objective function. The numerical results demonstrate that the proposed method accurately localize the user.
무선 주파수 신호와 단말 내장 센서를 이용한 심층 신경망 기반 융합 측위 기법
이현욱(Hyunwook Lee),임채훈(Chaehun Im),장수환(Suhwan Jang),이충용(Chungyong Lee) 대한전자공학회 2021 대한전자공학회 학술대회 Vol.2021 No.6
We propose an enhanced pedestrian localization based on a deep neural network. The input of DNN comprises estimated positioning parameters from built in sensors of mobile devices and radio frequency signal, such as step information, heading, and distances between device and access points (APs). By adopting DNN, the limitations of existing probability-based method, computation complexity and noise variance knowledge requirement, can be releaved. Moreover, simulation results show that the proposed localization method is more precise than the comparative method.