Mobile robot relocation using ambient magnetic fields and radio sources

In this paper we present a method for solving an indoor relocation problem using ambient magnetic fields and radio sources. Specifically, we exploit the magnetic field anomalies and noisy radio ranges in indoor environments. A robot with two magnetometers and one active radio range sensor first explores the unknown environment based on the SLAM technique, gathering its path and some useful multisensory observations. The gathered data is used to train a multilayer perceptron either in batch or stochastically. The trained model is then directly used to obtain the desired 2-D robot position when the first location is needed. The performance of the proposed method is validated by the experimental results.