利用测距值优化的室内三维定位算法

为了解决实际误差分布与假设误差分布不匹配导致最小二乘三维定位算法精度较低的问题,提出一种基于测距值优化的最小二乘室内三维定位算法.结合残差修正、Cayley-Menger行列式及空间四面体几何约束,利用非线性规划方法优化测距值,使测距误差符合高斯分布特性,从而提高最小二乘三维定位算法的定位精度.实验对比分析结果显示,所提算法具有较高的定位精度及稳定性.

A 3-D Indoor Localization Algorithm Using Distance Optimization

Least Square is a typical three-dimensional location algorithm for time of arrival based indoor positioning system.The precondition of traditional LS algorithm is that the measurement error meets the zero mean and the equal variance.However,the multipath and non-line of sight significant in realistic indoor environment leads TOA ranging error to be non-Gaussian distribution and cannot meet the hypothesis.This conflict leads low localization accuracy.This article presents a distance optimization based least square 3-D indoor localization algorithm.The nonlinear programming method with Cayley-Menger determinant and tetrahedral geometry constraints was adopted by the proposed algorithm to optimize measured distance,which makes the ranging error fit Gaussian distribution and thus improves the localization accuracy of least square.Experiments demonstrate that the proposed distance optimization based least square 3-D localization algorithm achieves better localization accuracy and stability.