Development of map-aided GPS algorithms for vehicle navigation in urban canyons

A major portion of the Location-Based Services (LBS) market deals with applications involving in-car navigation systems. The Global Positioning System (GPS) is the most popular choice for positioning in such applications. Many LBS applications involve positioning in urban areas having high rise buildings. Although GPS has good positioning accuracy in open sky conditions, it suffers from line-of-sight issues in urban canyons.; This thesis formulates some new methods for aiding GPS using maps for vehicle navigation in urban canyons. GPS satellite availability in urban canyons can be improved by using a High Sensitivity GPS (HS GPS) receiver which can track weak signals. However, this introduces large errors and noise in measurements. Thus, reliability monitoring becomes necessary with such receivers in signal degraded environments. Maps and Digital Elevation Models (DEM) provide effective constraints to compute an outlier-free solution. In this research, a robust fuzzy logic-based approach is developed for road segment identification. This identified road segment is then used in a GPS computation model and is referred to as Map Aided GPS (MAGPS). The performances of different map matching approaches are analyzed and results show that the proposed algorithms can be effectively used to navigate the vehicle in urban canyons as compared to the conventional GPS-based approaches. The input to the system comes from a low cost gyro (Murata ENV-05), a HS GPS receiver (SiRF XTrac) and a map database. The effect of height aiding from Digital Elevation Model (DEM) is also analyzed.