Reliability considerations of multi-sensor multi-network pedestrian navigation

Abstract
This study discusses a simple multi-sensor multi-network positioning system that integrates global positioning satellite (GPS) measurements, accelerometers and a digital compass with wireless network localisation utilising a pedestrian motion model and dead reckoning. The feasibility of the multi-technology system for seamless outdoor to indoor pedestrian navigation is discussed with the emphasis on reliability issues and adaptability requirements. The multi-sensor multi-network positioning system is developed for challenging navigation environments such as indoors and deep urban canyons. This study considers how to estimate and improve such a multi-sensor multi-network system's reliability and estimate its accuracy. An outdoor to indoor pedestrian test is conducted. Adaptive filtering performance of the multi-technology solution as well as general measurement quality monitoring and error detection when an over-determined solution is at hand is shown. Reliability estimation utilising adaptation in the form of environment detection to estimate the final positioning accuracy is also presented.

This publication has 13 references indexed in Scilit: