摘要
目前,基于RSSI(received signal strength indication)的指纹定位算法由于低成本、易实施的特性,逐渐成为室内定位技术的研究热点。然而,基于RSSI的WiFi指纹定位受到指纹点观测质量的影响,RSSI抖动较大时引起定位精度较低。考虑到GPR(Gaussian process regression)模型能够有效地平滑时间序列信号,提出了基于GPR模型的WiFi指纹定位改进算法。实验结果表明,该算法能够有效提高定位精度,定位精度可达到1m,点位误差在小于1.5m限差时,其可靠度可达到83.3%。
Nowadays fingerprint localization algorithm based on RSSI,has become the main indoor positioning technology owing to its low cost and easy implementation.However,the quality of the fingerprint point observation affect the accuracy of WiFi fingerprint localization algorithm based on RSSI,so the RSSI jitter will lower the accuracy of the positioning.Considering that GPR model can effectively smooth time series,this paper presents an improved WiFi fingerprint positioning algorithm based on GPR model.Experimental results show that the algorithm can greatly improve the positioning accuracy,and the positioning accuracy can reach to 1m.Moreover,the positioning error of the all points less than1.5m,the feasibility can reach to 83.3%.
作者
刘少伟
花向红
邱卫宁
张伟
贺小星
LIU Shaowei HUA Xianghong QIU Weining ZHANG Wei HE Xiaoxing(School of Geodesy and Geomatics, Wuhan University, Wuhan 430079,China Collaborative Innovation Center of Geospatial Technology, Wuhan 430079, China Hazard Monitoring ~ Prevention Research Center, Wuhan University, Wuhan 430079,China Jiangxi Province Key Laboratory for Digital Land, Nanchang 330013, China)
出处
《测绘地理信息》
2017年第5期46-49,共4页
Journal of Geomatics
基金
国家自然科学基金资助项目(41174010
41374011)
江西省数字国土重点实验室开放研究基金资助项目(DLLJ201605)
关键词
RSSI
WiFi指纹定位
指纹库
GPR预测模型
received signal strength indication
WiFi finger print positioning
fingerprint library
GPR prediction model
