Differential GPS - Understanding GPS: Principles and Applications - page 394

Table 8.1 Pseudorange Error Budget for GPS SPS With and Without LADGPS Corrections
Segment Source Error Source 1 Error (m)

GPS Only with LADGPS

Space/control Broadcast clock 1.1 0.0

L1 P(Y)-L1 C/A
group delay

0.3 0.0

Broadcast ephemeris 0.8

0.1-0.6 mm/km × baseline in km

User Ionospheric delay 7.0

0.2-4 cm/km × baseline in km

Tropospheric delay 0.2 1-4 cm/km × baseline in km

Receiver noise and
resolution

0.1 0.1

Multipath 0.2 0.3

System UERE Total (RSS) 7.1 0.3m + 1-6 cm/km × baseline in km

8.3.2 Regional-Area DGPS
To extend the region over which LADGPS corrections can be used without the
decorrelation of errors that accompanies the separation of the user from the station,
three or more reference stations may be distributed along the perimeter of the region
of coverage in a concept referred to as regional-area DGPS. The user receiver can
then obtain a more accurate position solution by employing a weighted average of
pseudorange corrections from the stations. Because the error in the broadcast cor-
rections grows with distance from each station, the weights may be determined by
geometric considerations alone to give the largest weight to the closest station, such
as by choosing those weights that describe the user position as the weighted sum of
the station positions [10]. For example, with three stations at locations denoted by
latitude

φ and longitude λ , the three weights, w 1 , w 2 , and w 3 , of stations M 1 ( φ 1 , λ 1 ),

M 2 ( φ 2 , λ 2 ), and M 3 ( φ 3 , λ 3 ) for user U ( φ , λ ) may be determined by the following set of

three equations (Figure 8.9):

φ ww w

ww w
ww w

= φ + φ + φ

11 2 2 3 3

λ = + +

++ =

λ λ λ

11 2 2 3 3

( )

1

1 2 3

A two-step approach to using multiple monitoring stations to improve the accu-
racy of the user's position estimate is described in [10]. In the first step, the
pseudorange corrections from each monitor are used to determine the position of
the user individually. The second step entails computing a weighted average of the
individual position estimates to provide a more accurate estimate. Each weight is
formed from the inverse of the product of the distance of the monitor from the user
and the standard deviation from the average of the estimates from that station, nor-
malized by the sum of the weights. The error introduced by each monitor receiver is
thus diluted by its weight, so that if, for example, the weights were all equal, then
each monitor receiver error would be diluted by a factor of 1/ n . But since the errors
are uncorrelated, the standard deviation of their sum is 1/ n ; thus, the standard