Sequential learning for fingerprint based indoor localization

详细信息    查看全文

• 作者：Nhan Vo Than Ngo ^{nvthanhan@gmail.com} ; Jeong Geun Kim ; ^{jg_kim@khu.ac.kr}
• 关键词：Indoor localization ; Convex optimization ; Fingerprinting
• 刊名：AEU - International Journal of Electronics and Communications
• 出版年：2017
• 出版时间：January 2017
• 年：2017
• 卷：71
• 期：Complete
• 页码：105-109
• 全文大小：581 K
• 卷排序：71

文摘

In this paper, we introduce a novel approach for improving performance of fingerprinting based indoor localization. Our proposal is a two-step procedure in which severe variation in the received signal strength is minimized during the first step via convex optimization, and distance metric learning is then used to estimate a more accurate location. Numerical results show that our proposal outperforms existing techniques in terms of accuracy and reliability.