Analog neural network approach for source localization using time-of-arrival measurements

Abstract

Source localization can be achieved by making use of the time-of-arrival (TOA) measurements, but it is not a trivial task because the TOAs have nonlinear relationships with the source coordinates. This paper exploits a neural network technique, namely, Lagrange programming neural networks, for TOA-based localization. We also investigate the local stability of our formulation. Simulation results demonstrate that the performance of the proposed location estimator approaches the optimality benchmark of Cramér-Rao lower bound. © 2012 Springer-Verlag.