Well “need” may be too big a word but there are very significant benefits of using more than one system. For high accuracy applications, which are our primary focus, there is a significant accuracy benefit. This benefit is especially visible and important in real time usage. The additional satellites of another GNSS do improve the real-time position estimates dramatically.
To explain this we have done a simple experiment. We have taken a static station, a station that does not move, but estimated the position for each 30 second epoch over a 24 hour period. The repeatability of the epoch estimates gives us a very good measure for the precision of the epoch estimates. Since the station does not move the estimates should be the same for each epoch, well at least if our model for the Earth tides and such is accurate.
To show the effect of GPS versus GNSS we selected a station that does track both the GPS and the GLONASS satellites. Thus we could generate 3 solutions, a GLONASS-only solution, a GPS-only solution, and a GNSS solution (GPS + GLONASS). The results are given in the bar plot below where the red bars represent the GLONASS solution, green the GPS solution, and blue the GNSS solution. And we distinguished between the East, North and Up component as Up is typically the worst component.
The bar plot nicely demonstrates that the GNSS solution (blue) is significantly better then the GPS solution (green), in the East and North directions almost by a factor of two. This is despite the fact that the GLONASS only solution (red) does not perform very well. The quality of the kinematic GNSS position estimates is astonishing with repeatability’s of 5 mm for the horizontal and 15 mm for the vertical components.
The relatively poor performance of the GLONASS solution is caused by the fact that the GLONASS constellation is not complete. Does during the 24 hour period we used for this experiment there is a period where there are less than 4 satellites in view, this severely impacts the quality of the GLONASS only solution but has no negative effect at all on the GNSS solution. This highlights another benefit of GNSS solutions; the larger number of satellites leads to much more reliable and robust estimates. This explains the main reason why especially surveyors and the offshore industry are fully “addicted” to GNSS. Besides being better it is much more robust and reliable.
Last but not least the picture above shows the 30 second position estimates of the GNSS solution. A repeatability of 5 mm for the horizontal and 15 mm for the vertical components for fully kinematic position estimates is quite impressive, is it not!? So regarding the question "to GPS or to GNSS" the answer is a clear GNSS.
