Date of Award
12-2009
Document Type
Thesis
Degree Name
Master of Science (MS)
Legacy Department
Computer Engineering
Committee Chair/Advisor
Hoover, Adam W.
Committee Member
Gowdy , John N.
Committee Member
Brooks , Richard R.
Abstract
A filter designed to be optimal for alternating multi-sensor tracking systems is presented. This filter can be used on ultra-wide band position tracking systems such as the Ubisense system. In this thesis, a comprehensive evaluation of the accuracy of the Ubisense tracking system in a multi-room building is given. Then the multi-modality of the noise distribution of such a system is shown. A multi-sensor tracking system is then simulated and a Kalman filter is used to filter the measurements. It is shown that the Kalman filter is not optimal for such a system and the need for a filter that addresses the issue of multi-modality is explained. Finally a modified particle filter is designed and is shown to effectively reduce the noise in an alternating multi-sensor position tracking system.
Recommended Citation
Ganjali, Danyan, "Filtering Noise Caused by Sensor Selection for an Ultra-wide Band Position Tracking System" (2009). All Theses. 717.
https://tigerprints.clemson.edu/all_theses/717