In order to eliminate the effects of indoor obstacles lead to ultra-wideband（ UWB） ranging result

present an UWB indoor localization algorithm based on Kalman filtering. By using signal noise ratio of UWB to distinguish line of sight from non-line of sight environment

give least squares calibration models for ultra-wideband sensor ranging performance to reduce the distance measuring system error. To weaken multipath effects and effect of non-line of sight error on ranging

use Kalman filtering priori estimation to replace posterior estimation to process ranging result. Extended Kalman filter is used to achieve indoor positioning. Experimental results prove that

the algorithm has sub-meter high dynamic realtime positioning precision in complex indoor environments.