Comparison of Filtering Methods for the Peer-to-Peer Kalman Filter

Isabelle Kraemer and Bernd Eissfeller

Abstract:	The popularity of so called smart phones which provide also a GPS or even A-GPS (Assisted-GPS) unit for navigation is still increasing. The main field of navigation covered by these devices is pedestrian and also indoor navigation. Although A-GPS (Assisted GPS) enables positioning also in light indoor areas it is still not possible to estimate a position when residing in deep indoors. The approach introduced in this paper is based on the work of [1] and [2]. Additionally to assisted GNSS (Global Navigation Satellite Systems), WiFi-assisted positioning strategies and the use of mobile radio network infrastructure for navigation the users’ devices (peers) profit of the knowledge of surrounding devices. To improve the position estimation of all peers within an ad-hoc network a filtering algorithm is used which computes a new position based on the estimated position of all peers and the quality of their position (variance). While in [1] and [2] a very simple averaging filter was used which already showed some promising results the approach was enhanced and an algorithm for Kalman Filtering was developed. Since the movement of people cannot be modeled by a linear function a Kalman Filter is necessary which is able to work with non-linear functions. Therefore one algorithm is implemented based on the equations of the Extended Kalman Filter (EKF) and two other algorithms based on the equations of the Unscented Kalman Filter. All three filter variations are implemented for Matlab and are tested on their functionality and performance using a simulation with a varying amount of peers. The results of the filtering techniques are compared with each other and serve as recommendation for the implementation of a prototype on a mobile device.
Published in:	Proceedings of the 24th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2011) September 20 - 23, 2011 Oregon Convention Center, Portland, Oregon Portland, OR
Pages:	2896 - 2905
Cite this article:	Kraemer, Isabelle, Eissfeller, Bernd, "Comparison of Filtering Methods for the Peer-to-Peer Kalman Filter," Proceedings of the 24th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2011), Portland, OR, September 2011, pp. 2896-2905.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In