
17
GP2000 – GPS CHIPSET DESIGNER’S GUIDE
The nominal carrier DCO setting is 88540000/63 or
1450396825Hz.
Formulation of Integrated Carrier Phase
Integrated carrier phase or integrated Doppler is used to
smooth the code phase pseudo-ranges:
fD =
2n3L1
c
where
fD is the Doppler frequency and n the relative velocity.
The integral of the Doppler frequency gives a measure of the
range change over the integration interval:
fD = 2
n3L1
c
fD = fDCO2fnom
where
fDCO is the carrier DCO frequency (Doppler shifted)
and
fnom is the nominal carrier DCO frequency (unshifted).
Therefore:
fDCO2 fnom = 2
n3L1
c
Nk2T3 fnom = 2
L1DR
c
where
Nk is the carrier DCO cycle count between TICs, T is the
TIC interval and
DR is the change in range between TICs.
Hence,
DR =
A running sum of
DR can then be used to smooth the code
pseudo-ranges.
∫
c (T3 fnom 2Nk)
L1
Determination of the Navigation Solution
The basic radial range measurement to satellite i is given
by:
= (x2x
i )
21(y2y
i )
21(z2z
i )
21ct = R
i
where
(x,y,z) is the receiver location, (xi,yi,zi) is the satel-
lite position,
t is the receiver clock offset and Ri the pseudo-
range.
To solve for the 4 unknowns,
(x,y,z) and t, 4 equations or
pseudo-ranges to 4 satellites are required. Commonly, a
linearised version of the range equation is used:
DRi =
where
(xn,yn,zn) and tn are the nominal (best estimates) of
(x,y,z) and t; Dx, Dy, Dz and Dt are the corrections to
these estimates;
Rni is the nominal pseudo-range
measurement and
DRi is the difference between the actual
and nominal measurement.
The range equations can then be expressed in matrix notation by:
Ax = r or x = A
21 r
where
A is the solution or design matrix (the direction cosine
matrix),
x is the receiver position and clock correction vector
and
r is the pseudo-range measurement difference vector.
(x2xi )
(Rni2ctn)
Dx1
Dx1cDt
(z2zi )
(Rni2ctn)
Dx1
(y2yi )
(Rni2ctn)
More generally, for the overdetermined case (more than 4
satellites), the method of least squares is used:
Ax2r = v
where
v is the vector of residuals. Since the solution matrix is
no longer square the generalised or pseudo-inverse must be
used:
x = (A
TWA)21ATWr
where
W is the weight matrix.
It can be shown that the optimum value of
W is the inverse
covariance matrix of the pseudo- ranges. This can be estimated
from the satellite URAs (User Range Accuracy) transmitted in
the satellite data messages.
If the solution is underdetermined (only 3 measurements),
or if the GDOP is above the desired GDOP mask, then altitude
aiding is used by adding an extra measurement for an imaginary
satellite at the centre of the Earth. The added pseudo-range is
equal to the magnitude of the current receiver range vector from
the Earth’s centre.
Pseudo-range rates are used in the same manner as pseudo-
ranges to determine the receiver velocity vector and clock drift.