White-LED 3-D indoor positioning using super-resolution technique

Abstract

This paper presents a 3-D indoor positioning algorithm using white light-emitting diodes (W-LED) in the presence of additive white Gaussian noise (AWGN). Adopting a uniform linear array (ULA) based photo-detector, the angle of irradiance and incidence of W-LEDs and photo-detector can be estimated by super-resolution techniques. Through Lambertian model, the corresponding distances can also be calculated. By forming geometric circles, 3-D target location can be determined via non-linear optimization of least-square (LS) criterion. To evaluate the efficacy of the proposed algorithm, MATLAB simulation is conducted for a space of 5m × 5m × 3m with three W-LED sources and one photo-detector device. The simulation result showed less than 10cm location inaccuracy which is a 2.5 fold improvement compared to other 3-D indoor positioning schemes recently presented in the literature.