that the pseudorange corrections are 
calculated in real-time and sent to user 
nearby via a communication link in order to 
eliminate most of the common errors in the 
conventional GPS system. 
Let the C/A code pseudorange measured 
by a GPS receiver at time t be written as 
Pr = Pr +S,u + Ap/ (1) 
where p 1 r represents the measured C/A code 
pseudorange between the base (reference) 
receiver r and satellite i, pi is the 
geometric range which is calculated from the 
receiver coordinates {x r ,y r ,z r ) and the 
satellite coordinates (x l ,y l ,z l ) at epoch t, 
8 tu is the clock offset of receiver r, and 
Ap r ‘ is the total delay error from receiver r 
to satellite i. From (1), the pseudorange 
correction A p r ‘ is found as: 
A p l r =pl- P l -8 tu (2) 
When the differential correction (2) is 
applied to the measured range of the remote 
GPS receiver, we get: 
Pi+Ap‘=pi (3) 
where p' m represents the measured C/A 
code pseudorange between the m-th 
receiver and satellite i, and p„ represents 
the corrected pseudorange. 
2.2 Code Pseudorange Smoothing 
Algorithm using Carrier Phase 
Measurements 
Generally speaking, when low-cost GPS 
receivers such as Garmin 100/75 are used, a 
positioning accuracy about 10 meter can be 
obtained when using the conventional DGPS 
algorithm of Section 2.1 in real-time (He, 
1996). This accuracy seems to be low for 
high-accuracy vehicle tracking and 
navigation systems. Hence, it is of interest to 
develop a more accurate approach to 
improve differential positioning accuracy. 
For DGPS positioning, the raw data is C/A 
code pseudorange. Hence, the accuracy is 
directly related to the positioning accuracy. 
The measurement error of code pseudorange 
is one of the main error sources in differential 
positioning. To improve the positioning 
accuracy of a vehicle, the measurement error 
of the pseudorange should be reduced. 
The measured pseudorange can be improved 
by using: