Kernel information for hyperbolic localization

Wan, Q.,Yang, W.L.,Ming Hui Wang,Peng, Y.N.
DOI: https://doi.org/10.1109/VETECF.2005.1559046
2005-01-01
Abstract:In terrestrial cellular wireless communication surroundings, the geometrical relationship between the measurements and the source position, from which the ordinary hyperbolic location algorithm is derived, is often destroyed by the unknown environmental effect of non-line of sight (NLOS) propagation. In this paper, the NLOS effect on mobile localization is first modeled in feature space, which is produced by a nonlinear mapping on both the original range-difference measurements and an ordinary hyperbolic localization estimate. Then a novel kernel information based hyperbolic localization method is proposed to extract the environment kernel information and correct the location bias of the ordinary location method. Simulation results demonstrate the applicability of the kernel information for hyperbolic localization.
What problem does this paper attempt to address?