An improved NLOS detection scheme using stochastic characteristics for indoor localization

Indoor localization scheme using sensor networks is expected to be applied in various fields, and the localization scheme using time of arrival (TOA) is well-known. However, the estimation accuracy of TOA localization is severely deteriorated in non-line-of-sight (NLOS) environments, and the NLOS mitigation scheme such as iterative minimum residual (IMR) scheme is required. The IMR scheme is often applied because of its lower calculation complexity. However, when an increased number of NLOS nodes exist, the NLOS detection errors increase in the IMR scheme and the estimation accuracy deteriorates. Therefore, in this paper, we propose a new scheme exploiting rough NLOS detection based on stochastic characteristics before the application of IMR scheme to improve the localization accuracy. The improved performance is shown by computer simulations.