Maximum Correntropy Criterion With Variable Center for NLOS Mitigation in Elliptic Positioning

This correspondence revisits the problem of robust elliptic positioning (EP) in the possible presence of unreliable bistatic range (BR) measurements and focuses specifically on mitigating the negative effects of the non-line-of-sight (NLOS) contaminated BR observations in EP. A major limitation of the existing robust statistics based EP solutions is that many of them assumed the fitting errors have a zero center, which may not be the case in NLOS environments tending to positively bias the BR data. We provide a practical remedy for this, by applying instead the maximum correntropy criterion with variable center (MCC-VC) to robustify the EP estimator and developing an alternating minimization algorithm for optimization. Simulations are conducted to demonstrate the superiority of our proposal over a number of off-the-shelf EP approaches. © 2024 IEEE.