Robust NLOS Error Mitigation Method for TOA-Based Localization via Second-Order Cone Relaxation

The existence of non-line-of-sight (NLOS) errors can significantly degrade the localization performance. In this letter, we address the time-of-arrival (TOA) based localization problem under NLOS conditions. To accurately localize a source, we propose a novel robust second-order cone relaxation (SOCR) method, which is not sensitive to NLOS errors. The proposed method does not require one to know any statistics of the NLOS errors, and instead, it only requires to know the upper bound on the magnitudes of the NLOS errors, which is relatively easy to obtain in practice. Simulation results show that the proposed robust SOCR method has good performance in various NLOS scenarios, and also it has a relatively lower computational complexity than the existing methods.