GLONASS, a satellite based navigation system similar to GPS, has a great po
tential for precise positioning. However, unlike GPS, GLONASS satellites tr
ansmit signals at different frequencies, which results in complexity in pro
cessing GLONASS carrier phase measurement. The standard double-difference p
rocedure used in precise GPS positioning cannot be used in precise GLONASS
positioning. In this paper, a mathematical proof for this problem is given.
The five existing modelling methods are then theoretically reviewed. A zer
o baseline data st is processed using these methods to show their performan
ces. The paper concludes with a comparison of these modelling methods.