TY - GEN
T1 - Computer vision based pre-processing for channel sensing in non-stationary environment
AU - Gao, Wei
AU - Peng, Wei
AU - Liu, Jiajia
AU - Nie, Zhifeng
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/9
Y1 - 2019/9
N2 - With the evolution of wireless networks, new techniques including massive multiple-input multiple- output (MIMO) and millimeter wave are adopted to satisfy the demands for diversified services. However, it has been verified by field tests that the traditional wide sense stationary assumption for wireless channel does not hold anymore. As a result, traditional channel state information (CSI) acquisition methods, especially the statistical CSI acquisition, cannot be applied straightforwardly in such a circumstance. In this paper, we propose a pre-processing method for channel sensing in the non-stationary environment. Specifically, the data sampled from channel training is treated as a channel image, where the statistical channel state is represented by gray-scale. Then the computer vision technique, specifically, the edge detection method, is used on the channel image to detect the homogeneous sub-regions. Within each sub-region, the channel is statistically stationary, and then the CSI can be obtained by existing methods. It is verified by simulation results that, the proposed method can help to improve the CSI acquisition accuracy in the non- stationary environment.
AB - With the evolution of wireless networks, new techniques including massive multiple-input multiple- output (MIMO) and millimeter wave are adopted to satisfy the demands for diversified services. However, it has been verified by field tests that the traditional wide sense stationary assumption for wireless channel does not hold anymore. As a result, traditional channel state information (CSI) acquisition methods, especially the statistical CSI acquisition, cannot be applied straightforwardly in such a circumstance. In this paper, we propose a pre-processing method for channel sensing in the non-stationary environment. Specifically, the data sampled from channel training is treated as a channel image, where the statistical channel state is represented by gray-scale. Then the computer vision technique, specifically, the edge detection method, is used on the channel image to detect the homogeneous sub-regions. Within each sub-region, the channel is statistically stationary, and then the CSI can be obtained by existing methods. It is verified by simulation results that, the proposed method can help to improve the CSI acquisition accuracy in the non- stationary environment.
UR - http://www.scopus.com/inward/record.url?scp=85075255608&partnerID=8YFLogxK
U2 - 10.1109/VTCFall.2019.8891457
DO - 10.1109/VTCFall.2019.8891457
M3 - 会议稿件
AN - SCOPUS:85075255608
T3 - IEEE Vehicular Technology Conference
BT - 2019 IEEE 90th Vehicular Technology Conference, VTC 2019 Fall - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 90th IEEE Vehicular Technology Conference, VTC 2019 Fall
Y2 - 22 September 2019 through 25 September 2019
ER -