Comparative Performance Analysis of a Kalman Filter and a Modified Double Exponential Filter for GPS-Only Position Estimation of Automotive Platforms in an Urban-Canyon Environment

This paper compares the performance of a modified double exponential filter (DEF) with respect to a Kalman filter for automotive applications that rely solely on position estimation via a global-positioning-system receiver. The performance of these two filters was initially analyzed in simulations for both Gaussian and non-Gaussian noise sources on the position estimates. The analysis was extended to include a field data collection on a vehicle in an urban-canyon environment. The modified DEF exhibited better 2-D position accuracy in both the non-Gaussian simulation and the field data collection efforts.