A new automated cycle slip detection and repair method
approach may provide unexpected results (Dai et al. 2009).
As a matter of fact, at present there is only one modernized
GPS satellite in space broadcasting signals at L1, L2, and L5
three frequencies. The use of dual-frequency GPS receivers
and satellites is still prevailing in the applications.
In this paper, a new cycle slip detection and repair method
that employs the ionospherictotal electron content (TEC) rate
(TECR) is proposed. Unlike other methods where the iono-
spheric variation is regarded as a nuisance or simply ignored,
in this method, the ionospheric variation, characterized by
the TEC rate, is precisely estimated and used to detect and
repair cycle slip. Without cycle slips, the ionospheric phys-
ical TECR is normally bounded by a certain value. In this
study, we choose the threshold value as 0.15 TECU/s. When
there are cycle slips, the ionospheric TECR will become
significantly larger, with most contribution from the artifi-
cial ionospheric TECR that is resulted from cycle slips. The
Melbourne–Wübbena wide lane (MWWL) ambiguity will
also change when there are cycle slips. With the data of both
ionospheric TECR change and MWWL ambiguity change,
the cycle slips can be precisely and uniquely determined.
This paper is organized as below. The methodology of
detecting and repairing cycle slip is developed in detail in
Sect. 2. The accuracies of the estimated cycle slip terms are
analyzed too. The analyses of the method performance are
performed in Sect. 3. In Sect. 4, the application of this method
is discussed. The conclusion is given in Sect. 5.
2 Methodology
In this cycle slip detection and repair method, two sets of
measurements are jointly employed to precisely determine
the size of the cycle slips. One is the Melbourne–Wübbena
linear combination that has been widely used for cycle slip
detection and ambiguity resolution because of its 86-cm long
wavelength (Melbourne 1985; Wübbena 1985; Blewitt 1990;
Kass et al. 2009; Bock et al. 2009). This combination removes
the effects from the atmosphere (including both ionosphere
and troposphere), the geometry, and the satellite and receiver
clocks. Therefore, this combination is useful to check GPS
observations for cycle slips (Dach et al. 2006). The other set
of measurements is the ionospheric TECR that is inferred
from L1 and L2 carrier phase measurements.
2.1 Carrier phase and pseudorange observations
We assume that the dual-frequency GPS receiver observes
the satellite p. The observation equations for dual-frequency
GPS carrier phase and pseudorange measurements can be
written as:
λ
1
p
1
= ρ
p
+ c(dt − dT
p
) − I + T + λ
1
N
p
1
, (1)
λ
2
p
2
= ρ
p
+ c(dt − dT
p
) − γ I + T + λ
2
N
p
2
, (2)
P
p
1
= ρ
p
+ c(dt − dT
p
) + I + T, (3)
P
p
2
= ρ
p
+ c(dt − dT
p
) + γ I + T, (4)
where λ
1
and λ
2
are the wavelengths of the GPS L1 and
L2 signals, respectively;
p
1
and
p
2
are carrier phase mea-
surements on L1 and L2 frequencies, respectively; P
p
1
and
P
p
2
are the pseudorange measurements; ρ
p
is the geometrical
distance between the receiver and the satellite p; dt and dT
p
are the GPS receiver and satellite clock errors, respectively;
c is the speed of light in vacuum; I is the ionospheric range
delay on GPS L1 signal; T is the tropospheric range delay;
N
p
1
and N
p
2
are the integer number of cycles for GPS L1 and
L2 signals, respectively, which are often called ambiguities;
γ = f
2
1
/ f
2
2
is the ratio of the squared frequencies of GPS
L1 and L2 signals. The carrier phase measurements
p
1
and
p
2
and the ambiguities N
p
1
and N
p
2
are in the unit of cycle.
The dt and dT
p
are in the unit of second. The rest variables
are all in the unit of meter.
2.2 Cycle slip determined from the Melbourne–Wübbena
wide-lane combination
The well-known MWWL linear combination at a given epoch
can be formed as below (Melbourne 1985; Wübbena 1985).
The tag for epoch is omitted for brevity.
L
MWWL
=
f
1
· λ
1
p
1
− f
2
· λ
2
p
2
f
1
− f
2
−
f
1
· P
p
1
+ f
2
· P
p
2
(
f
1
+ f
2
)
= λ
WL
N
p
WL
, (5)
where λ
WL
= c/
(
f
1
− f
2
)
≈ 86 cm and N
p
WL
= N
p
1
− N
p
2
.
The λ
WL
and N
p
WL
are called wide-lanewavelength and wide-
lane ambiguity, respectively. Thus, the wide-lane ambiguity
at a given epoch can be estimated as:
N
p
WL
=
L
MWWL
λ
WL
=
p
1
−
p
2
−
f
1
· P
p
1
+ f
2
· P
p
2
λ
WL
(
f
1
+ f
2
)
. (6)
We assume that at epoch (k−1) there are no cycle slips, or the
cycleslips havebeen repaired if any.But at epoch (k) there are
cycle slips on both L1 and L2 carrier phase measurements
p
1
(k) and
p
2
(k). The cycle slips are denoted as N
p
1
(k)
and N
p
2
(k), respectively. After cycle slips are detected and
repaired, the correct carrier phase measurements at epoch (k)
should be calculated as:
p
1
(k) =
p
1
(k) + N
p
1
(k), (7)
p
2
(k) =
p
2
(k) + N
p
2
(k), (8)
where
p
1
and
p
2
denote the correct L1 and L2 phase mea-
surements, respectively, after cycle slips are repaired.
123