Date of Award

8-2012

Embargo Period

10-11-2012

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Electrical and Computer Engineering

Advisor(s)

Dan Stancil

Second Advisor

Tamal Mukherjee

Abstract

Pedestrian navigation is readily enabled by GPS in outdoor environments. However, there are many locations - such as indoors, urban canyons and underground - where the GPS signal is unreliable or unavailable. Compact inertial navigation, through the use of MEMS accelerometers and gyroscopes, has been applied to a pedestrian’s shoes as a means to navigate without GPS. MEMS inertial technology has issues with sensor bias drift, causing significant navigational errors. Simulations by Brand and Phillips and experimental results from Laverne et al. have shown that the addition of a scalar range between shoes can constrain this error growth. A shoe ranging sensor (SRS) using radar technology, due to its relative insensitivity to environmental factors such as dirt and dust, will be applied to provide this scalar range. Commercially available radars generally operate over longer ranges (>> 1 m) and relatively course accuracy (>1 cm). This dissertation proposes stepped frequency continuous wave (SFCW) modulation will fill the technology gap and provide a ranging solution with <1 cm RMS accuracy in a 1 m range. The SRS architecture is extended using time division multiplexing to generate multiple ranges for estimating relative shoe heading. Theoretical limits on SFCW and combined CW ranging accuracy are derived. A ray-tracing channel model is developed for SRS simulation to analyze ranging performance on different surfaces, antennas and ranging algorithms (Fourier and Prony estimation). The channel model extends the Friis transmission equation to include propagation phase and develops a method to calculate phase from arbitrary antenna polarization and orientation. Measurements performed validate channel model and demonstrate antenna limited sub-cm ranging performance on metal, concrete and anechoic foam. Walking tests of the SRS co-mounted with an acoustic range sensor show good agreement.

Share

COinS