Assisted GNSS - Traditional and Vectorized: Implementation and Performance Results

This paper outlines the current situation and problems for making use of a GPS-enabled device in urban areas. Rather than increasing acquisition sensitivity by further leveraging computational expensive long integration, the paper examines the positioning performance by merging both GPS and GLONASS observations to achieve higher availability and accuracy. To make it more comparatively intuitive, three assisted "push-to-fix" GPS/GLONASS algorithm implementations are compared - they are: (a) traditional A-GPS algorithm implementation; (b) AGPS/GLONASS algorithm implementation; (c) AGPS/GLONASS algorithm implementation leveraging a vectorized approach. The final algorithm leverages all acquired signals from both constellations to promote sensitivity and/or suppress the probability of false alarm. All the implementations are tested with a consecutive live dataset collected in downtown Denver, CO, USA. The experimental results reveal that in challenging urban environment, "more measurements adoption" significantly benefits the assisted solution. Additionally, the proposed vectorized technique is verified to advance the reliability of the position determination by combining correlation powers from all available satellites.Remote SensingEICPCI-S(ISTP)