Characterizing Multipath from Satellite Navigation Measurements in Urban Environments

This paper focuses on the characterization of multipath in satellite navigation measurements for urban environments using a modern receiver. In order to increase the accuracy of the Global Positioning System (GPS) in urban environments for the purpose of automotive navigation, the impact of multipath should be mitigated. However, to better understand the operational environment and the algorithms best suited for this, such locations must first be characterized, which is the purpose of this research. This goal was accomplished by collecting GPS data in urban environments such as downtown San Francisco using two different receivers. Range and position data were recorded for each receiver in order to characterize multipath in both the range and position domain. Raw Intermediate Frequency (IF) data were also collected and post processed using a reconfigurable software-based GPS receiver architecture, allowing a more thorough analysis of the multipath induced distortion modes of the correlation function used in tracking the signal. Using methods identified in various literature, the levels of multipath error are assessed for individual pseudor-anges in urban environments and compared to the level of multipath in suburban and rural environments. The impact of multipath induced range error is projected into the position domain with respect to the geometric dilution of precision (GDOP) and satellite availability. The impact of short term biases due to multipath on the position solution is also evaluated.