Abstract: The localization accuracy could be severely degraded due to the non-line-of-sight (NLOS) propagation in the indoor environment. A NLOS signal detection algorithm based on hypothesis testing was proposed, and a new particle swarm optimization algorithm based minimum residual localization algorithm was presented for the wireless sensor networks. The received signal strength and time-of-arrival (TOA) were used to distinguish between LOS and NLOS propagation based on hypothesis testing. Simulation results showed that compared with deviation analysis and fewer parameters of propagation model, this method has lower computation complexity. Finally, particle swarm optimization algorithm based minimum residual localization algorithm was proposed after propagation detection algorithm. The proposed method outperforms the existing weighted least square method and Fang method with better estimation accuracy.