基于位置指纹的混合定位算法的研究

摘要

With the popularity of mobile intelligent terminals and the rapid development of mobile Internet, loca-tion-based services and applications appeared in large numbers. Low-cost and high-precision indoor positioning solution has become a new hotspot. Location fingerprinting and dead reckoning are two main techniques to solve indoor positioning problem. Location fingerprinting treats received signal strengths as recognizable patterns, trains classification model with pattern recognition algorithms and determine location based on the input received signal strengths. Dead reckoning esti-mates next position by speed and direction information base on the previous estimate. This paper proposes a hybrid algo-rithm based on location fingerprinting and dead reckoning which enhances the accuracy. Furthermore, this paper proposes an online model update algorithm which can adapt to the slow changes of radio-map. This algorithm can reduce labor costs on repeated fingerprint collections.%随着移动智能终端的普及和移动互联网的迅猛发展，基于位置的服务和应用大量涌现。低成本、高精度的室内定位解决方案已成为新的研究热点。位置指纹和航位推算是解决室内定位问题的两个主要技术。位置指纹法将接收信号强度作为可识别的模式，用模式识别的方法训练出一个分类模型，根据接收信号强度信息判别定位者所在的位置。航位推算法将前一时刻的位置估计作为起点，利用速度和方向信息估计定位者下一时刻的位置。本文以位置指纹法和航位推算法为原型，提出了一种混合定位算法，该算法可以有效提高定位精度。此外，本文提出了一种在线更新定位模型方案，能适应无线环境的缓慢变化，节省重复采集指纹的人力成本。