Home > Published Issues > 2016 > Volume 4, No. 2, December 2016 >

The NLOS Localization Algorithm Based on the Linear Regression Model of Extended Kalman Filter

Nan Hu, Chengdong Wu, Li Chen, and Hao Wu
Northeastern University, Shenyang, Liaoning, China

Abstract—The location of the mobile node is an important issue in wireless sensor network (WSN). In WSN area, the NLOS (Non-line-of-sight) propagation of signal is ubiquitous and has a significant influence on the localization accuracy. Based on the Linear Regression Model of Extended Kalman Filter (EKF), the beacon node status is identified and the distance residuals is produced in this paper. Then the H-Infinity filter algorithm is used to filter the NLOS distance measurement values. Finally the maximum likelihood localization method is used to locate the position. Simulation results demonstrate that the proposed algorithm have a higher localization accuracy than other methods in different environments and have strong robustness in terms of inhibit NLOS errors.

Index Terms—wireless sensor network, mobile localization, non line of sight, extended Kalman filter, H-infinity filter

Cite: Nan Hu, Chengdong Wu, Li Chen, and Hao Wu, "The NLOS Localization Algorithm Based on the Linear Regression Model of Extended Kalman Filter," Journal of Image and Graphics, Vol. 4, No. 2, pp. 141-144, December 2016. doi: 10.18178/joig.4.2.141-144