Experimental Measurement and Characterization of Ionospheric and Multipath Errors in Differential GPS

Real-time differential GPS is fundamentally limited by a) signal estimation errors, b) satellite ephemeris errors, c) propagation decorrelation errors, and d) local multipath errors. In this paper we address the latter two error sources. An understanding of these is needed to establish an error budget for any monitor station network that the U.S. Coast Guard would consider for a nationwide system. To estimate these error sources, extensive two-frequency P-code and integrated Doppler data have been collected over 150 kilometer and 1500 kilometer baselines. A processing scheme is developed which employs both code and carrier observations. Applicable to real-time differential systems, as well as to post-collection data analyses, the estimation scheme treats multipath errors and carrier-code offsets as state-variables. The resulting Kalman filter yields the very precise carrier observation, corrected for carrier-code offset. Day-to-day repeatability of multipath errors is illustrated. The decorrelation of ionospheric errors across short and long baselines is evaluated by combining the integrated Doppler estimation scheme with the standard two-frequency ionospheric measurement technique. Also discussed are the impact of multipath and ionospheric error sources on Differential GPS.