基于多目标优化的AP选取新算法

A new AP selection algorithm based on multi-objective optimization

讨论基于RSS的Wi-Fi室内定位中基于多目标优化的AP选取问题。综合考虑信息增益(JIG)和互信息(MI)的多目标优化函数的推导过程,同时利用基因算法(GA)寻求多目标函数的最优解。两种差异明显的环境下的实验结果表明:复杂环境中信息增益和互信息最佳权重分别为0.3和0.7,稳定环境中互信息的最佳权重分别为0.7和0.3,同时位置估计结果分析表明:不同AP个数下的位置误差的方差和平均值之间存在明显的线性相关性。

This paper explores a new access point （AP） selection algorithm based on multi-objective optimization for Wi-Fi indoor localization. A derivative process of multi-objective optimization function which involves both joint information gain （JIG） and mutual information （MI） is described in detail. The genetic algorithm （GA） is used to find the optimal solution. Experiments are conducted under different environments and the localization results suggest that the best weights of joint information gain and mutual information are 0. 3 and 0. 7 respectively in the complex environment while in the stable environment the best weights of joint information gain and mutual information are 0. 7 and 0. 3. And the results also show a linear relationship between the variance of error values of different AP numbers and the average value.