Comparative analysis of the standard error in relative GNSS positioning for short, medium and long baselines.

Abstract

: In GNSS Relative positioning, both in post processing and real time positioning, GNSS baseline processing plays a vital role in determining the standard error in the positioning. This paper presents a comparative analysis of the standard error in relative GNSS positioning for short, medium and long baselines. Satellite observations were acquired on 19 control points within the campus of university of Lagos, using both the passive and active (CORS) station principles in Post Processing GNSS positioning data at differential mode. The short baselines have maximum length not exceeding 1.5km from the control points to a base station in university of Lagos, the medium baselines have range not exceeding 12km from the control points to a CORS located at the Lagos state Surveyor General’s Office, Nigeria, while the long baselines have range not exceeding 107km from the control points to another CORS (Continuously Operating Reference Station) located in Cotonuo, Benin Republic. After post processing operation, the standard error in relative position was computed from the satellite geometry model. The results were statistically analysed using ANOVA One Way at 0.05 significant level comparing the average standard error in relative positioning of all stations during the three baseline observation scenario. Subsequently, Scheffe test was conducted on the ANOVA results. The statistical results show no significant difference between the level of standard error obtained by the baseline processing involving the CORS at 12km and the Base receivers at 1.5km but there exists a significant difference between the 107km CORS baseline processing and the 12km as well as the 1.5km baselines. The short baselines were found to have the highest achievable processing precision while the long baselines have the least. The study shows that the longer the baselines the lower the standard error in relative positioning even with CORS. This however, does not negate the reliability of the long baseline results but defines the level of precision and accuracy achievable when compared with other baseline length with a view to further more researches on long baseline error reduction.