Robust Indoor Localization on a Commercial Smart-Phone - Robotics Institute Carnegie Mellon University

Robust Indoor Localization on a Commercial Smart-Phone

Nisarg Kothari, Balajee Kannan, and M. Bernardine Dias
Tech. Report, CMU-RI-TR-11-27, Robotics Institute, Carnegie Mellon University, August, 2011

Abstract

This technical report outlines a system for robust indoor localization on a commercial smart-phone. Towards achieving effective localization, in the absence of GPS, we combine complementary localization algorithms of dead reckoning and Wifi signal strength fingerprinting. Dead reckoning is performed using the on-board accelerometer, magnetometer, and gyroscope sensors to detect motion and estimate orientation. At the same time, Wifi signal strength fingerprinting is employed to provide an independent position estimate. These measurements along with a pre-built map of the environment are combined using a particle filter towards more accurate pose estimation. We discuss a procedure for collecting Wifi calibration data which uses a robot to reduce the amount of time needed to train the system for a given environment. The system was tested using multiple participants in two different indoor environments. It was found to achieve localization accuracies of within 10 meters, sufficient for a variety of navigation and context-aware applications.

BibTeX

@techreport{Kothari-2011-7342,
author = {Nisarg Kothari and Balajee Kannan and M. Bernardine Dias},
title = {Robust Indoor Localization on a Commercial Smart-Phone},
year = {2011},
month = {August},
institute = {Carnegie Mellon University},
address = {Pittsburgh, PA},
number = {CMU-RI-TR-11-27},
keywords = {GPS-free localization, RSSI fingerprinting, dead-reckoning solutions, indoor positioning system, Particle filters},
}