A Mobile Pedestrian Dead Reckoning System

Abstract

A pedestrian dead reckoning system estimates a pedestrian's position relative to a known location, and has a variety of applications. It is commonly used with a GPS receiver; however, the GPS signals may be affected by many factors and consequently unable to position correctly. This paper presents a pedestrian dead reckoning system that includes an inertial module, a GPS receiver module, and a dead reckoning module. The inertial module detects the acceleration of a step and heading. Step length is difficult to determine since it is affected by many factors and varies from person to person. Accurate stride estimation is significant in a pedestrian dead reckoning system. When the GPS is able to position, the Particle Swarm Optimization algorithm is applied to discover a pedestrian's step length coefficient. The obtained coefficient and the detected acceleration signal are used to determine the stride of a pedestrian. From the information of step length and heading, the dead reckoning can be processed without the GPS reception. The experimental results have demonstrated that the system presents good performances in either walking distance estimation or walking trajectory determination without knowing a pedestrian's leg length or assistance of wireless sensor network.