NLOS Identification and Correction Based on Multidimensional Scaling and Quasi-Accurate Detection

In wireless sensor networks, most of the previous NLOS identification is based on error estimation model established by raw data. In this paper, we propose a method of NLOS identification named NIMQ based on multidimensional scaling (MDS) and Quasi-Accurate detection (QUAD). In this method, we first map NLOS information into gross error information by MDS, then we use QUAD to identify the gross errors which contain the NLOS information. This method relies only on distance measurements and is independent of the measured error estimation model. In addition, using the network topology constraints in higher dimensional space, the identified distance can be corrected by multiple iterations. Finally, an NLOS iterating correction algorithm (NICA) is proposed. Simulations show that in different scenarios our proposed NIMQ and NICA can well identify and correct NLOS measurement.