基于物理邻近点辅助的无线局域网指纹定位方法

发布时间：2018-06-14 22:28

本文选题：无线局域网定位 + 指纹库　；参考：《计算机应用》2014年06期

【摘要】：针对传统位置指纹图中的邻近参考点(ARP)信息未能得到较好利用的问题,提出一种在离线训练阶段建立基于接收信号强度(RSS)的位置指纹库和参考点(RP)物理邻近信息库的方法。通过利用待定位点与其所对应的最近邻参考点及参考点之间的物理邻近关系,来提高指纹概率定位方法的定位精度,即:在在线定位阶段,首先根据基于信号强度概率分布的贝叶斯算法计算得到待定位点的最近邻点;然后在物理邻近信息库中搜索最近邻点的物理邻近点,并在该最近邻和物理邻近点集合中,选取特征点集合用于贝叶斯二次估计;最后将具有最大后验(MAP)概率的特征点组的均值位置作为待定位点的估计位置。实验结果表明,与传统的无物理邻近数据库的指纹概率定位方法相比,在3 m内的定位精度提高了约10%,有效提高了定位的可靠性。
[Abstract]:Aiming at the problem that the ARP) information of the traditional position fingerprint map is not well utilized, a method of establishing the location fingerprint database and the physical proximity information base based on the received signal strength (RSS) in the off-line training stage is proposed. By using the physical proximity relationship between the undetermined site and its nearest neighbor reference point, the location accuracy of the fingerprint probabilistic location method is improved, that is, in the online location stage, First, the nearest neighbor of the undetermined site is calculated according to Bayesian algorithm based on the probability distribution of signal intensity, and then the physical adjacent point of the nearest neighbor is searched in the physical proximity information base, and in the set of the nearest neighbor and the physical adjacent point, The set of feature points is selected for Bayesian quadratic estimation. Finally, the mean position of feature point group with maximum posterior map probability is used as the estimated position of undetermined site. The experimental results show that, compared with the traditional fingerprint probabilistic location method without physical proximity database, the positioning accuracy within 3 m is improved by about 10%, and the reliability of the location is improved effectively.
【作者单位】：移动通信技术重庆市重点实验室(重庆邮电大学);
【基金】：国家自然科学基金资助项目(61301126) 国家科技重大专项(2012ZX03006-002(3)) 重庆市基础与前沿研究计划项目(cstc2013jcyjA40041,cstc2013jcyjA40032,cstc2013jcyjA40034) 重庆市教委科学技术研究项目(KJ130528) 长江学者和创新团队发展计划项目(IRT1299) 重庆邮电大学博士启动基金资助项目(A2012-33) 重庆邮电大学青年科学研究项目(A2012-77)
【分类号】：TN925.93

【参考文献】