Bayesian navigation system with particle filtering and dead reckoning in urban canyon environments

A Bayesian navigation system is designed to enhance the performance of the position estimate for a pedestrian in urban canyon environments using particle filtering and digital compass on Smartphone. The particle filtering is used to correct the inaccurate GPS positioning data (position and direction) caused by the strong multipath in the urban areas. The digital compass readings are employed to correct the directional measurements of the positioning information estimated by the particle filtering, but they need to be calibrated with another particle filtering and GPS since they are subject to cumulative errors. In this demo, we demonstrate a real implementation of the proposed Bayesian navigation system on the iPhone 4S. The demo shows the enhanced performance of the new navigation scheme compared to using only GPS signals.