一种基于主成分回归的DV-Hop定位方法

DV-Hop定位算法是一种被广泛运用的定位算法。在各向同性的密集网络中,DV-Hop可以得到比较合理的定位精度,然而在实际分布的网络中,它的精度受到噪声和信标节点之间几何关系的限制。主成分回归方法利用主成分分析方法对原先数据进行重新构造,删除部分主成分,从而消除部分噪声和多重共线性对回归精度、稳定性的影响。根据DV-Hop算法定位过程,在节点位置估计阶段运用主成分回归的方法对定位数据进行重新综合与提取,仅利用有效定位信息进行位置估计。仿真实验结果证明该改进后的算法同样具有原先算法优良特性,且定位精确度有所提高。

A DV-Hop localization algorithm using principal component regression

DV-Hop algorithm is one of the important localization algorithms and is widely applied to various positioning system.The algorithm performs better in isotropic density sensor networks.However,the localization accuracy is limited by noise and the relative geometry of the beacon nodes.The principal component regression method uses principal component analysis method to reconstruct the original data,and by deleting some of the principal components from the regression,the effects of parts of noises and multi-collinearity on the accuracy and stability of the regression can be eliminated.According to the localization principle of the DV-Hop algorithm,an improved DV-Hop localization method is proposed.The principal component analysis is used to re-integrated and extraction localization data and only effective location information is used for location estimation in the estimation phase of node coordinates.Simulation results show that the improved algorithm inherits the excellent characteristics of the original algorithm with an improvement on the location accuracy.