RSS
指纹(计算)
欧几里德距离
计算机科学
到达角
相似性(几何)
聚类分析
多输入多输出
人工智能
模式识别(心理学)
指纹识别
算法
计算机视觉
频道(广播)
电信
天线(收音机)
图像(数学)
操作系统
作者
Chen Wei,Kui Xu,Zhexian Shen,Xiaochen Xia,Weixin Xie,Lihua Chen,Jinming Xu
标识
DOI:10.1109/iccc51575.2020.9344979
摘要
Fingerprint based localization is an effective positioning method for rich scattering environments, which has attracted the enormous attention in recent years. In this paper, we propose a novel fingerprint positioning method for cell-free massive multiple-input multiple-output (MIMO) systems. The angle-domain channel power matrix with lots of angle information can be extracted as the arrival-of-angle (AOA) fingerprint by exploiting discrete Fourier transform (DFT) operation. Then we also propose the angle similarity coefficient and the Euclidean distance as the AOA and received signal strength (RSS) fingerprint similarity criterions respectively to evaluate the distance between two fingerprints. Moreover, the K-means clustering algorithm is performed for improving the efficiency of fingerprint matching. Finally, we utilize the weighted K-nearest neighbor (WKNN) algotithm to estimate the location of the user, whose weight can be constructed according to the above fingerprint similarity criterions. The simulation results demonstrate that our proposed joint AOA-RSS fingerprint based location method has the better positioning performance than the methods only consider AOA or RSS fingerprint.
科研通智能强力驱动
Strongly Powered by AbleSci AI