基于二进制传感器网络的目标运动方向估计方法

A Method for Estimating the Moving Direction of Target based on Binary Sensor Network

介绍了方向估计二进制传感器网络(DE-BSN),提出基于统计学习解决目标运动方向估计的思路,并给出了使用线性判别分析法(LDA)和支持向量机算法(SVM)的估计方案。仿真分析了在不同传感器节点密度时上述算法的精度、和在节点出现错误的情况下算法对目标方向估计的可靠性。实验结果表明上述算法均可实现高精度的目标方向估计,并都具有一定的鲁棒性,各自的优势分别在于:LDA的计算复杂度较小,而SVM的估计误差较小。

A target direction evaluation model with binary sensor network （DE-BSN） is presented in this paper. We implemented two algorithms of statistical learning theory,namely Linear Discriminative Analysis （LDA） and Support Vector Machine（SVM）. A performance study of accuracy of them with a variety of sensor node density was conducted through simulation. We further simulate the robustness of them with different error node density. The results suggest that both algorithms perform high accuracy and moderate robustness. LDA has lower computation complexity,and SVM perform low error rate.