Li et al.

Satellite Navigation (2022) 3:7

https://doi.org/10.1186/s43020-022-00068-0
Satellite Navigation
https://satellite-navigation.springeropen.com/

ORIGINAL ARTICLE Open Access

Principle and performance

of multi‑frequency and multi‑GNSS PPP‑RTK
Xingxing Li1, Bo Wang1, Xin Li1*, Jiaxin Huang2, Hongbo Lyu1 and Xinjuan Han1

Abstract
PPP-RTK which takes full advantages of both Real-Time Kinematic (RTK) and Precise Point Positioning (PPP), is able
to provide centimeter-level positioning accuracy with rapid integer Ambiguity Resolution (AR). In recent years, with
the development of BeiDou Navigation Satellite System (BDS) and Galileo navigation satellite system (Galileo) as well
as the modernization of Global Positioning System (GPS) and GLObal NAvigation Satellite System (GLONASS), more
than 140 Global Navigation Satellite System (GNSS) satellites are available. Particularly, the new-generation GNSS
satellites are capable of transmitting signals on three or more frequencies. Multi-GNSS and multi-frequency observa-
tions become available and can be used to enhance the performance of PPP-RTK. In this contribution, we develop
a multi-GNSS and multi-frequency PPP-RTK model, which uses all the available GNSS observations, and comprehen-
sively evaluate its performance in urban environments from the perspectives of positioning accuracy, convergence
and fixing percentage. In this method, the precise atmospheric corrections are derived from the multi-frequency and
multi-GNSS observations of a regional network, and then disseminated to users to achieve PPP rapid AR. Furthermore,
a cascade ambiguity fixing strategy using Extra‐Wide‐Lane (EWL), Wide-Lane (WL) and L1 ambiguities is employed to
improve the performance of ambiguity fixing in the urban environments. Vehicle experiments in different scenarios
such as suburbs, overpasses, and tunnels are conducted to validate the proposed method. In suburbs, an accuracy of
within 2 cm in the horizontal direction and 4 cm in the vertical direction, with the fixing percentage of 93.7% can be
achieved. Compared to the GPS-only solution, the positioning accuracy is improved by 87.6%. In urban environments
where signals are interrupted frequently, a fast ambiguity re-fixing can be achieved within 5 s. Moreover, multi-
frequency GNSS signals can further improve the positioning performance of PPP-RTK, particularly in the case of small
amount of observations. These results demonstrate that the multi-frequency and multi-GNSS PPP-RTK is a promising
tool for supporting precise vehicle navigation.
Keywords: Multi-frequency, Multi-GNSS, PPP-RTK, Rapid ambiguity resolution, Vehicle navigation

Introduction positioning technologies, is able to achieve centime-

With the development of emerging technologies such ter-level positioning accuracy globally using a single
as autonomous driving, mass-market applications receiver (Zumberge et al., 1997; Kouba and Héroux
are in urgent need of fast, high-precision, and low- 2001). However, it needs several minutes to initialize,
bandwidth location services (European GNSS Agency, which inhibits its application in the mass-market (Geng
2019). Precise Point Positioning (PPP), as one of the et al., 2020; Li et al., 2018, 2019). Recently, a PPP-RTK
mainstream Global Navigation Satellite System (GNSS) method that exploits the precise atmospheric correc-
tions from a regional network to enable rapid Ambi-
guity Resolution (AR) was proposed (Li et al., 2013;
*Correspondence: xinli@sgg.whu.edu.cn Teunissen et al., 2010; Wubbena et al., 2005). The PPP-
1
School of Geodesy and Geomatics, Wuhan University, 129 Luoyu Road, RTK technique exhibits unique advantages in terms of
Wuhan 430079, China convergence, accuracy, and real-time communication,
Full list of author information is available at the end of the article

© The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.
Li et al. Satellite Navigation (2022) 3:7 Page 2 of 11

which has become a forefront subject in the field of In this contribution, we developed a multi-frequency
vehicle navigation. and multi-GNSS PPP-RTK model, which uses all the
The concept of PPP-RTK was firstly proposed by available GNSS observations, to achieve rapid position-
Wubbena et al. (2005), aiming to achieve centim- ing with the accuracy at centimeter-level for vehicle
eter-level positioning accuracy in a few seconds in navigation in urban environments. In addition, several
post-processing and real-time applications with the vehicle experiments were conducted in different scenar-
augmentations from regional networks. Later, Teunis- ios such as suburbs, overpasses, and tunnels to compre-
sen et al. (2010), Li et al. (2011), and Zhang et al. (2011) hensively verify the effectiveness of the proposed model.
established the prototypes of the PPP-RTK system The benefits of multi-frequency and multi-GNSS obser-
respectively, which commonly have two parts: on the vations used for PPP-RTK are analyzed in terms of the
server side, the precise orbit, clock, phase bias as well accuracy, convergence, and fixing percentage. This study
as the atmospheric corrections are generated based is organized as follows: Sect. 2 introduces the algorithm
on the observations of a regional network; on the user of the multi-frequency and multi-GNSS PPP-RTK system
side, fast AR of absolute positioning is realized with the in detail. Section 3 provides the experimental procedures
corrections from the network. The atmospheric mod- and the data processing strategies in the vehicle experi-
eling and bias correcting in PPP-RTK were first studied ments. Section 4 evaluates the performance of the multi-
by academia (Li et al, 2014; Oliveira et al., 2017; Zhang frequency and multi-GNSS PPP-RTK and analyzes its
et al, 2019)., and then PPP-RTK method has come into results in different scenarios. Section 5 gives conclusions
practical applications. The Japanese Quasi-Zenith Sat- and outlooks.
ellite System (QZSS) is the first system to implement a
satellite-based PPP-RTK augmentation service—Cen- Method
timeter Level Augmentation Service (CLAS) (Cabinet Multi‑frequency and multi‑GNSS PPP‑RTK system
Office, 2020). In addition, some commercial companies The procedure of the multi-frequency and multi-GNSS
have also provided PPP-RTK services, e.g., Trimble PPP-RTK system is illustrated in Fig. 1. On the server
CenterPoint RTX-FAST, NovAtel TerraStar-X, Ublox side, given precise satellite orbits and clocks, PPP is con-
PointPerfect, etc.. However, most of the services adopt ducted at all stations of the regional network. Then, the
dual-frequency observations, and their performances satellite phase biases, also known as the Uncalibrated
in real-time kinematic positioning need improving. Phase Delays (UPDs), of Extra-Wide-Lane (EWL), Wide-
In recent years, with the development of BeiDou Lane (WL), and Narrow-Lane (NL) are estimated by the
Navigation Satellite System (BDS) and Galileo naviga- method proposed by Li, Han, et al. (2021). Therewith, the
tion satellite system (Galileo) as well as the moderni- multi-frequency and multi-GNSS PPP AR is carried out
zation of Global Positioning System (GPS) and GLObal to extract the precise atmospheric corrections. On the
NAvigation Satellite System (GLONASS) (Li et al., user side, the atmospheric corrections of the user side are
2019; Montenbruck et al., 2014; Ren et al., 2021), more interpolated using the atmospheric corrections from the
than 140 GNSS satellites are available and new-genera- server side. When the satellite orbits and clocks, UPDs,
tion GNSS satellites are capable of transmitting signals
on three or more frequencies. It has been demonstrated
that the convergence, accuracy and reliability of PPP
AR can be significantly improved by multi-GNSS fusion Observations
Observations
of reference Orbit and clock
(Jokinen et al., 2013; Liu et al., 2017). The Time to First stations
of users

Fix (TTFF) can be shortened to 10 min when the obser-

vations of GPS/GLONASS/Galileo/BDS were adopted Server
PPP PPP
(Li et al., 2018). At the same time, the studies show that
the performance of PPP AR can be effectively improved
when multi-frequency observations are adopted (Geng Ambiguity
Estimating
EWL-WL-L1
& Bock, 2013; Gu et al., 2015; Li et al., 2018). Li et al. EWL/WL/NL
resolution cascade AR
UPD
(2019) found that the triple-frequency Galileo/BDS-2 User
(BeiDou-2 Navigation Satellite System) PPP AR can Extracting Interpolation PPP-RTK
converge to centimeter-level positioning accuracy in atmospheric
solution
corrections
about 10 min. Multi-frequency and multi-GNSS obser-
vations bring more possibilities for enhancing the per- Fig. 1 Flow chart of multi-frequency and multi-GNSS PPP-RTK system
formance of PPP-RTK.
Li et al. Satellite Navigation (2022) 3:7 Page 3 of 11

and the external atmospheric corrections are received, effects, Sagnac effect, solid earth tide, polar tides, ocean
PPP and EWL-WL-L1 cascade AR are implemented, and tide loadings, satellite and receiver antenna Phase Center
finally the multi-frequency and multi-GNSS PPP-RTK Offsets (PCOs) and Phase Center Variations (PCVs),
is achieved. In the following sections, the algorithms for and phase wind-up are corrected by the known models
atmospheric correction extraction based on the uncom- (Kouba, 2015).
bined PPP model on the server side, and for EWL-WL-L1 The uncombined PPP model is implemented to derive
cascade AR with external atmospheric corrections on the the precise tropospheric and ionospheric delays directly
user side are described. when ambiguity is fixed (Zhang et al. 2012). The gen-
eral equations of the multi-frequency undifferenced PPP
Atmospheric correction extraction based on uncombined model can be written as:
PPP AR model  s s
The GNSS pseudorange observation ( Pr,n
s ) and phase
s
p = µsr · x + t̂r − t IF + γi · Îr,i +msr,w Zr,w + er,k
s
 r,i


s
observation(Lsr,n) are described as: s s s s s
pr,j = µr · x + t̂r − t IF + γj · Îr,i +mr,w Zr,w + er,k

 s s s
pr,k = µsr · x + t̂r − t IF + γk · Îr,i +msr,w Zr,w + BkIFB + er,k
s

s
Pr,k =ρrs + c(tr − t s ) + γk · Ir,1
s
+ msr ·
(2)
s
Zr + (br,k − bks ) + er,k
(1) s
 s
l = µsr · x + t̂r − t IF − γi · Îr,i s
+msr,w Zr,w + i N̂r,i
s s
+ εr,i
Lsr,k =ρrs + c(tr − t s ) − γk · Ir,1
s
+ msr ·  r,i


s s
lr,j s
= µsr · x + t̂r − t IF − γj · Îr,i +msr,w Zr,w + j N̂r,j
s s
+ εr,j
Zr + k (Br,k − Bks ) + k · Nr,k
s s
+ εr,k 
 s s
lr,k = µsr · x + t̂r − t IF − γk · Îr,is
+msr,w Zr,w + k N̂r,ks s

+ εr,k
where the superscripts s, r and k represent satellite, (3)
receiver, and frequency, respectively; ρrs is the geometric With

fj 2

f2
k = i, k = j, αij = 2 i 2 , βij = 2


fi − fj2

fi − fj





 � �
s
t IF = c · t s + αij bis + βij bjs







 � �
t̂ c t α b β b

= · + +
 r r


 ij r,i ij r,j

 � �
 Î s = I s + β b − b � − β bs − bs
 �
ij r,i r,j ij
r,i r,i i j (4)

 IFB
�� s
� � s
�� βij �� s
� � s
��

 B k = br,k − bk − b r,i − b i + b r,i − b i − br,j − b j
βik




 � � ��
s s
+ Br,i − Bis + bIF s
− br,IF + βij br,i − br,j − βij bis − bjs /i
 � � � �
N̂r,i = Nr,i





 � � � � ��

s s s s s s
� �
N̂ N B B b b b b b b


 r,j
 = r,j + r,j − j + IF − r,IF + γj β ij r,i − r,j − γj β ij i − j /j


 � � ��
 N̂ s = N s + B − Bs + bs − b + γ β b − b − γ β bs − bs /
 � � � �
r,k r,k r,k k IF r,IF k ij r,i r,j k ij i j k

distance between the phase centers of the satellite where pr,j s

and lr,j
s
represent the observed-minus-com-
antenna and the receiving antenna; c is the vacuum speed puted values of pseudorange and phase observations,
of light; tr and t s denote the receiver and satellite clock respectively; µsr is the unit vector from the receiver to the
offset, respectively; Ir,1
s
refers to the ionospheric propaga- satellite; x ( x = [dx, dy, dz]) denotes the receiver posi-
tion delay of a GNSS signal on the first frequency; γk is tion increments vector relative to a prior position. BkIFB
the frequency-dependent multiplier factor, which can be is the Inter-Frequency Bias (IFB), which describes the
expressed as γk = k /1; msr · Zr represents tropospheric between-frequencies differences of code hardware delays.
path delay; br,k and bks are the code hardware delays while The code IFB at the frequency which is used for satellite
Br,k and Bks are the phase delays of receiver and satel- clock estimation can be absorbed into the ionospheric
lite, respectively; Nr,k
s represents the integer ambiguity; delay, while the others are absorbed by setting an individ-
er,k and εr,k are the sum of measurement noise and mul-
s s ual receiver clock parameter for each one. And the phase
tipath error of code and phase observations, respectively. IFB can be absorbed into ambiguity (Montenbruck, et al.
It should be noted that other errors such as relativistic 2010; Pan, Zhang, et al., 2017; Pan et al., 2019). Another
Li et al. Satellite Navigation (2022) 3:7 Page 4 of 11

bias that needs to be considered is the Inter-Frequency is employed to interpolate the precise atmospheric
Clock Bias (IFCB), which is induced due to the inconsist- corrections at the user side. As shown in Fig. 2, the
ency of the observations used for satellite clock estima- external ionospheric and tropospheric corrections are
tion and precise positioning. The Code-specific IFCB introduced into the multi-frequency and multi-GNSS
(CIFCB) is usually eliminated by Differential Code Bias undifferenced PPP as virtual observation equations.
(DCB) transformation (Guo et al., 2015), while the Phase Therefore, the PPP model with external atmospheric
corrections is expressed as:

s

s

 pr,n = µsr · x + t̂r − t IF + γn · Îr,i
s
+msr,w Zr,w + κ · BkIFB + er,n
s

 l s = µs · x + t̂r − t s − γn · Î s +ms Zr,w + n N̂ s + εs


r,n r IF r,i r,w r,n r,n
(5)
s s s 2
 l
 I
 = Îr,n − Ĩr,n = wI , wI ∼ N(0, σI )

 l s = Z − Z̃ = w , w ∼ N(0, σ 2 )

Z r,w r,w Z Z Z

IFCB (PIFCB) is corrected by the corresponding correc- with the stochastic model
tions (Pan, Li, et al., 2017; Pan, Zhang, et al., 2017). For
multi-GNSS PPP, a similar strategy with IFB is applied to � = diag(σp2 , σl2 , σI2 , σZ2 ) (6)
compensate the Inter-System Bias (ISB) which represents
where κ is equal to 0 or 1, Ĩr,n
s and Z̃
r,w denote the precise
the between-constellations differences of code hardware
ionospheric and tropospheric corrections, respectively;
delays (Khodabandeh et al. 2016). After the precise sat-
wI and wZ mean the differences between atmospheric
ellite orbits, clocks, and UPDs are received, PPP AR is
corrections and actual atmospheric delays, which are
conducted in the reference network. Then the precise
the white noise with zero-mean and variance of σI2 and
tropospheric delays and ionospheric delays are derived
σZ2 . After getting the PPP solution, the ambiguity at each
from the PPP AR solution directly, and broadcasted to
frequency is acquired and then transformed to EWL and
users as atmospheric corrections.
WL ambiguity according to Eq. (7).
EWL‑WL‑L1 cascade AR with external atmospheric
 s 
� s � � N̂r,i � s
�
corrections N̂EWL_jk dr,EWL − dEWL
�
0 1 −1 
 N̂ s  +

=
With the precise atmospheric corrections from a s
N̂WL_ij 1 −1 0  r,j  s
dr,WL − dWL
s
regional network, the Modified Linear Combina- N̂r,k
tion Method (MLCM) proposed by Li et al. (2011) (7)
where dr and ds are the UPDs of receiver and satellite,
respectively.
With the EWL, WL and NL UPDs products, the EWL,
WL and L1 ambiguities are fixed sequentially. The phase
Orbit and GNSS Atmospheric EWL/WL/NL biases of satellites are compensated with the correspond-
clock observations correction UPD
ing UPD products, and those of receivers are eliminated
through between-satellites differencing. Once the integer
UD float ambiguity UD PPP property is recovered, the ambiguity can be fixed by the
LAMBDA search method (Teunissen 2010), and intro-
Fixing EWL ambiguity duced to the PPP model as a virtual observation equation
EWL float ambiguity and update equation with an infinite weighting factor. The PPP model with the
external EWL AR constraint is written as:
Fixing WL ambiguity
WL float ambiguity and update equation L1UPD  s s s
p = µsr · x + t̂r − t IF + γik Îr,i +msr,w Zr,w + κ · BkIFB + er,n
s
 r,n


s s
Fixing L1ambiguity lr,n = µsr · x + t̂r − t IF − γik · Îr,i
s
+msr,w Zr,w + N̂r,n
s s
+ εr,n

 m = �N̂ s1 s2

=N s1 s2
+d s1 s2
r,EWL_jk r,EWL_jk EWL_jk
PPP-RTK solution (8)

Fig. 2 Flow chart of EWL-WL-L1 cascade AR with external

atmospheric corrections
Li et al. Satellite Navigation (2022) 3:7 Page 5 of 11

with stochastic model conducted in a suburban with an open-sky view, while

  the second one was in an urban area where the signals are
� = diag σp2 , σl2 , σm2 (9) frequently blocked by buildings, overpasses, and tunnels.
Both experiments last about half an hour. The road vehi-
σm2 is the variance of the virtual observation equation. The cle is equipped with a Novatel OEM729 GNSS receiver
constraint decorrelates ambiguity and other parameters, and a Trimble AT1675-540TS GNSS antenna to acquire
which can accelerate convergence and provide accurate multi-frequency and multi-GNSS data. In addition, a tac-
ambiguity. Similar to EWL AR, the WL ambiguity is fixed tical IMU is set in the vehicle to provide a reference tra-
and involved in the PPP model as a virtual observation jectory by processing PPK and Inertial Navigation System
equation which can be expressed as: (INS) with the data from the SBG IMU system.
s1 s2 s1 s2 s1 s2
Table 1 details the specific processing strategies used in
l = �N̂r,WL_ij = Nr,WL_ij + dr,WL_ij , σl2 ∈ 0 (10) the multi-frequency and multi-GNSS PPP-RTK system.
The orbit and clock errors are corrected by the prod-
With those constraints, the accuracy of the PPP solu- ucts provided by the Centre for Orbit Determination in
tion is improved and the search space of L1 ambiguity is Europe (CODE). The satellite phase biases are corrected
narrowed. After L1 UPD is obtained by UPD transforma- by the UPD products estimated by an open-source soft-
tion, L1 ambiguity is fixed. Finally, the multi-frequency ware called GREAT-UPD (https://​geode​sy.​noaa.​gov/​gps-​
and multi-GNSS PPP-RTK solution is derived. toolb​ox/) using the aforementioned 29 GNSS stations.
The UPDs and atmospheric corrections are updated
Experiments every 5 s, while the sampling interval of observations at
A vehicle experiment was conducted from 1:32 to 5:14 the user station is 1 s. The observations with elevation
GPS Time (GPST) on March 18, 2021 to evaluate the angle lower than 7° or the number of satellites in one
performance of the proposed multi-frequency and multi- epoch less than 4 are excluded.
GNSS PPP-RTK method. As shown in Fig. 3a, 29 GNSS
stations plotted by blue and red triangles in northern Results
China were used to estimate UPDs. 10 GNSS stations Evaluation of atmospheric corrections
(denoted by blue triangles in Fig. 3a) with an average To validate the atmospheric correction extraction
spacing of about 30 km were selected to provide precise method and the accuracy of the atmospheric correc-
atmospheric corrections. Figure 3b shows the trajecto- tions, stations A017, A018, and A019 from the Beijing
ries of the experimental vehicle. The first experiment was regional network are selected as the reference stations,

42.0°N
a b
41.5°N

41.0°N

40.5°N
Suburban
40.0°N Urban
39.5°N

39.0°N

38.5°N

38.0°N
114°E 115°E 116°E 117°E 118°E 119°E
Fig. 3 a Distribution of 29 stations and b the trajectories of the experimental vehicle
Li et al. Satellite Navigation (2022) 3:7 Page 6 of 11

Table 1 Processing strategy of multi-frequency and multi-GNSS PPP-RTK system

Items Strategies

GNSS system GPS, Galileo and BDS

GNSS signals GPS: L1, L2 and L5; Galileo: E1, E5a, E5b, E5 and E6; BDS: B1I, B2I and B3I
Combination mode Raw observations
Data sampling interval Server: 5 s; User: 1 s
Elevation mask 7°
Minimum number of satellites 4
Estimator Sequential least squares
Weight of observations Elevation-dependent weight
Phase wind-up effect Corrected
Satellite antenna phase center Corrected by igs14.atx
IFCB Corrected by IFCB products
Receiver antenna phase center Corrected by igs14.atx
Receiver clock Epoch-wise estimated for each system and each frequency
Phase ambiguities EWL-WL-L1 cascade partial fixing
Receiver coordinate Server side Fixed
User side Estimated in the epoch-wise kinematic
model
Ionospheric delays Server side Epoch-wise estimated for each satellite
User side Corrected by precise ionospheric cor-
rections
Tropospheric delays Dry component Modeled by Saastamoinen with Global Mapping function (GMF)
Wet component Server side Random-walk estimated
User side Corrected by precise tropospheric cor-
rections

and used to interpolate the atmospheric corrections of the atmospheric corrections. The difference results of
station A022 which is regarded as the user station. The zenith tropospheric corrections and slant ionospheric
distribution of these four stations is shown in Fig. 4. corrections for GPS, Galileo and BDS satellites are
Based on the method proposed in Sect. 2.2, the origi- shown in Fig. 5, with solid dots in each color represent-
nal atmospheric corrections of these four stations are ing one satellite in Panel 1–3. It is visible that most of
extracted. Then, the interpolated atmospheric correc- the differences are less than 1 cm. Some fluctuations are
tions of the user station are obtained according to the probably caused by the low elevation or short visible
MLCM mentioned in Sect. 2.3. The differences between time of the satellite. The Root Mean Square (RMS) of
the original and interpolated atmospheric corrections slant ionospheric differences for GPS, Galileo, and BDS
of station A022 are used to evaluate the accuracy of are 0.010, 0.008, and 0.006 m, respectively. Likewise,

40.3°N 0.1
Difference of atmospheric corrections (m)

0
A008 A017 −0.1
40.1°N A007 0.1
0
A025
Latitude

−0.1
39.9°N A013 A022 0.1
A014
0

A019 A018 −0.1

0.02
39.7°N
0
A003
−0.02
39.5°N 02:00:00 03:00:00 04:00:00 05:00:00 06:00:00
115.9°E 116.1°E 116.3°E 116.5°E 116.7°E GPS time
Longitude Fig. 5 Differences between the original and interpolated GPS, Galileo
Fig. 4 Distribution of stations A017, A018, A019, and A022 and BDS ionospheric corrections, and tropospheric corrections
Li et al. Satellite Navigation (2022) 3:7 Page 7 of 11

Table 2 The statistical result of the DF GPS-only, the DF GEC and the MF GEC PPP-RTK
PPP-RTK Solution RMS in different directions (m) Fixing percentage (%) TTFF (s)
E direction N direction U direction

DF GPS-only 0.134 0.159 0.280 68.60 1

DF GEC 0.035 0.048 0.099 87.80 1
MF GEC 0.019 0.017 0.035 93.70 1

Table 3 The statistical result of DF and MF GEC PPP-RTK

PPP-RTK RMS in different directions (m) Fixing TTFF (s) E direction N direction U direction
1
solution percentage
E N U 0
(%)

Positioning errors (m)

DF GPS
direction direction direction −1
E direction N direction U direction
1
GF GEC 0.217 0.217 0.346 62.66 1
0
MF GEC 0.040 0.052 0.134 83.94 1 −1 DF GEC
E direction N direction U direction
1
0
MF GEC
−1
E direction N direction U direction 04:08:20 04:20:00 04:31:40 04:43:20
Positioning errors (m)

5 GPS time
0 Fig. 7 Positioning series of DF GPS-only, DF GEC, and MF GEC
−5 PPP
PPP-RTK in the first experiment
E direction N direction U direction
1
0
−1 PPP-RTK
04:08:20 04:20:00 04:31:40 04:43:20 shows the DF GPS, Galileo, BDS (GEC) positioning
GPS time
series of PPP and PPP-RTK of the first experiment. The
Fig. 6 DF GEC positioning series of PPP and PPP-RTK in the first DF GEC PPP re-converges frequently due to the persis-
experiment
tent signal interference, and the re-convergence com-
monly takes more than ten minutes. In some situations,
the positioning accuracy of PPP even decreases to sev-
the differences series of zenith tropospheric correc-
eral meters. In contrast, the positioning series of DF
tions is quite stable with the RMS value of 0.004 m. Our
GEC PPP-RTK is quite stable with a fast convergence in
results indicate that the accuracy of the atmospheric
a few seconds. The RMS values of DF GEC PPP are 0.725,
corrections is at millimeter-level, which is sufficient for
1.005, and 1.467 m while those of DF GEC PPP-RTK are
users to achieve PPP AR.
0.035, 0.048, and 0.099 m in the east, north and vertical
components, respectively. Compared to the DF GEC PPP
Performance of multi‑frequency and multi‑GNSS PPP‑RTK
solution, the PPP-RTK solution is capable of shorten-
In order to comprehensively investigate the performance
ing the convergence time, and improving accuracy from
of the multi-frequency and multi-GNSS PPP-RTK, the
meter-level to centimeter-level.
results of PPP and PPP-RTK, single- and multi-GNSS
Figure 7 shows the positioning series of the DF GPS-
PPP-RTK, Dual-Frequency (DF) and Multi-Frequency
only, DF GEC, and MF GEC PPP-RTK in the first experi-
(MF) PPP-RTK, are compared, respectively. The per-
ment. A number of outliers can be seen from the series
formance of positioning solutions is evaluated in terms
of the DF GPS-only PPP-RTK, which indicates that the
of positioning accuracy, fixing percentage, and TTFF.
single-system PPP-RTK exhibits high instability in a
Here, the fixing percentage is calculated as the ratio of
kinematic environment. Compared to the DF GPS-only
the number of epochs with fixed solution to total num-
solution, the DF GEC PPP-RTK presents a more stable
ber of epochs. TTFF means the start time of a continuous
positioning series with fewer outliers, revealing a higher
5 s fixed solution whose positioning errors are less than
fixing percentage. Among all solutions, the MF GEC
5 cm in the horizontal direction and 10 cm in the vertical
PPP-RTK exhibits the best performance with the high-
direction.
est positioning accuracy and the least outliers. The RMS
The first experiment was conducted in an open-sky
values of the DF GPS-only PPP-RTK are 0.134, 0.159,
suburban where the signals were interrupted by road
and 0.280 m in the east, north and vertical components,
signs, trees, and low-rise buildings occasionally. Figure 6
Li et al. Satellite Navigation (2022) 3:7 Page 8 of 11

three solutions can converge to centimeter-level accuracy

20 10
in 1 s, but the GEC PPP-RTK shows the highest position-

PDOP
Positioning errors (m) NSAT

10 5
0 0
ing accuracy and fixing percentage.
1
E direction N direction U direction Figure 8 shows the number of the available GNSS sat-
0 ellites (NSAT) with precise atmospheric corrections and
DF GEC Position Dilution of Precision (PDOP) series and the
−1
E direction N direction U direction
1 positioning series of the DF and the MF GEC PPP-RTK
0 in the second experiment. In the urban environment, the
MF GEC
−1 NSAT generally decreases to 12. The GNSS signals are
02:20:00 02:27:55 02:35:50 02:43:45 02:51:40
GPS time seriously interrupted with high-rise buildings between
Fig. 8 NSAT and PDOP and positioning series of DF and MF GEC 02:45:40 and 02:46:40 GPST, and completely blocked by
PPP-RTK in the second experiment a tunnel between 02:33:52 and 02:34:38. Compared to
the DF GEC PPP-RTK, the MF GEC has a more precise
and stable positioning series with fewer outliers, par-
ticularly when NAST series are sharply declining. In the
respectively (Table 2). With the multi-GNSS observa- environment with the signal greatly interrupted, the MF
tions, the DF GEC PPP-RTK achieves the centimeter- GEC PPP-RTK can still achieve AR with the constraint of
level positioning accuracy of 0.035, 0.048, and 0.099 m EWL ambiguity, while GF cannot. The statistical results
in the east, north and vertical components, respectively. of positioning accuracy are given in Table 3. Compared
Compared to the DF GPS-only PPP-RTK, the corre- to the GF GEC PPP-RTK, the positioning accuracy of
sponding improvement is 73.88%, 69.81%, and 64.64%, the MF GEC PPP-RTK is improved from 0.217, 0.217,
respectively. Better yet, because of the addition of multi- and 0.346 m to 0.040, 0.052, and 0.134 m, exhibiting
frequency signals, the MF GEC PPP-RTK shows the the improvement of 81.57%, 76.04%, and 61.27% in the
highest positioning accuracy of 0.019, 0.017, and 0.035 m east, north, and vertical components. It can also be seen
in the east, north and vertical components. Compared to from Table 3 that the fixing percentage is improved from
the DF GPS-only and the DF GEC PPP-RTK, the posi- 62.66% to 83.94%. In general, the MF GEC PPP-RTK
tioning accuracy has improved by 87.58% and 62.49%. performs better and is able to meet the centimeter-level
Correspondingly, the fixing percentage is up to 93.70% accuracy needed for urban positioning.
which is increased by 25.10% and 5.90%, respectively. All

Table 4 The statistical result of DF and MF GEC PPP-RTK in Scenario (a)

PPP-RTK solution RMS in different directions (m) Fixing percentage TTFF (s) Re-convergence
(%) time (s)
E direction N direction U direction

GF GEC 0.097 0.036 0.410 80.00 1 22

MF GEC 0.033 0.011 0.094 96.67 1 1

Fig. 9 a Observational environments in the overpass plus buildings scenarios and b tunnel plus buildings scenarios
Li et al. Satellite Navigation (2022) 3:7 Page 9 of 11

likelihood, the loss lock of satellite and severe multipath

20 10
contribute to the wrong fixing of ambiguity (Zhang & Li,
NSAT

PDOP
10 5
0 0 2013), leading to outlies in the positioning series of the
60 DF solution between 02:46:13 and 02:46:26 GPST. Hence,
as shown in Table 4, it takes 22 s to re-converge to a cen-
SNR
(dB)

40
20 timeter-level accuracy after crossing the overpass. Finally
2 the RMS value in the vertical direction is 0.410 m and
Positioning Positioning Multipath
(m)

0
the fixing percentage is only 80.00%. Fortunately, with
−2
E direction N direction U direction the constraint of the EWL ambiguity, the MF GEC PPP-
errors (m) errors (m)

1
0
RTK maintains high-precision and stable positioning
−1
DF GEC throughout the experiment. Even though GNSS signals
E direction N direction U direction are interfered by buildings frequently, it realizes AR and
1
0 centimeter-level positioning accuracy in 1 s. Therefore, a
MF GEC
−1 reliable fixed solution with a fixing percentage of 96.67%,
02:45:40 02:45:50 02:46:00 02:46:10 02:46:20 02:46:30 02:46:40
GPS time and a higher positioning accuracy with the RMS values
Fig. 10 NSAT, PDOP, SNR, multipath and positioning series of DF and of 0.033, 0.011, and 0.094 m in the east, north and ver-
MF GEC PPP-RTK in Scenario (a) tical components, respectively, is successfully achieved.
Compared to the DF mode, the accuracy is improved by
76.31% and the fixing percentage is increased by 16.67%.
Multi‑frequency PPP‑RTK performance in different We speculate that the addition of multi-frequency obser-
scenarios vations can resist the influence of multipath.
Two scenarios, i.e., overpass plus buildings (Fig. 9a) and In the second scenario, the vehicle went downhill
tunnel plus buildings (Fig. 9b) were selected to validate starting at 02:32:20 GPST where the NSATs gradually
the benefits of MF signals for the GEC PPP-RTK. The decreased to 10 as shown in Fig. 11. During the period
observational environments for both scenarios are shown from 02:32:52 to 02:34:30 GPST, the vehicle was crossing
in Fig. 9. The experimental vehicle first passed through the tunnel where the signals were completely lost. After-
the overpass and then ran along the road with build- ward, the vehicle left the underground tunnel and the
ings on its sides in the first scenario, and in the second
scenario it passed a tunnel with dense buildings in the 20 // 10
NSAT

PDOP
entrance and exit of the tunnel. 10 5
In the first scenario, the experimental vehicle crossed 0 // 0
the overpass at 02:45:58 GPST for four seconds. Then, 60 //
SNR
(dB)

the vehicle entered a road with dense buildings on its 40

20 //
sides, where GNSS signals are seriously disturbed. The
2 //
positioning series as well as the NSAT, PDOP, Signal–
Positioning Positioning Multipath

0
(m)

Noise-Ratio (SNR), and code multipath of the observa- −2 //

tion on the first frequency of GPS, Galileo and BDS series 1 E direction N direction U direction
errors (m) errors (m)

are shown in Fig. 10. The NSAT varies mostly between 0

DF GEC
12 and 14, and the corresponding PDOP is smaller than −1 //
E direction N direction U direction
2. It is also found that when the NSAT drops below 10, 1

the PDOP value rises to above 2 between 02:45:58 and 0

MF GEC
−1 //
02:46:09 GPST. Most of the SNRs range from 40 to 60 dB, 02:32:20 02:32:30 02:32:40 02:32:50 02:34:40 02:34:50
but below 40 for the period 02:45:58 to 02:46:15 GPST. GPS time

The code multipath series behave like a random noise Fig. 11 NSAT, PDOP, SNR, multipath and positioning series of DF and
MF GEC PPP-RTK in Scenario (b)
series, where meter-level outliers exist from 02:46:00
to 02:46:20 GPST and 02:46:31 GPST to the end. In all

Table 5 The statistical result of the DF and MF GEC PPP-RTK in Scenario (b)
PPP-RTK solution RMS in different directions (m) Fixing percentage TTFF (s) Re-convergence
(%) time (s)
E direction N direction U direction

GF GEC 0.338 0.533 0.769 40.38 – 3

MF GEC 0.029 0.045 0.148 80.77 1 1
Li et al. Satellite Navigation (2022) 3:7 Page 10 of 11

signal intensity gradually increased. The majority of SNR Moreover, the performance of the proposed method in
series is between 40 and 60 dB, and the values of mul- the overpass and the tunnel scenarios was assessed. The
tipath effect are most in the range of ± 1.5 m. The posi- results show that accuracy of within 5 cm in the hori-
tioning series of the DF GEC PPP-RTK cannot converge zontal direction and within 15 cm in the vertical direc-
to centimeter-level accuracy until the vehicle leaving the tion, and the TTFF of 1–3 s can be achieved. The fixing
tunnel. In contrast, the MF GEC PPP-RTK maintains a percentage is 96.67% in the overpass scenario, while
more precise, stable and complete positioning results 80.77% in the tunnel scenario. Even though GNSS signals
when GNSS signals are available. It is worth noting that are interfered by the buildings or blocked by the tunnel,
there are still some places where the positioning results AR and centimeter-level positioning accuracy can be
are missing or have large errors due to the obscuration achieved in 3 s.
of the buildings. The statistical results of the two solu- The results of the above experiment indicate that the
tions are given in Table 5. Compared to the DF GEC PPP- multi-frequency signals can resist the influence of mul-
RTK, the MF improves accuracy from the RMS of 0.338, tipath on the PPP-RTK, and significantly improve the
0.533, and 0.769 m to RMS of 0.029, 0.045, and 0.148 m accuracy and reliability of the PPP-RTK in the urban
in the east, north and vertical components, respectively. environment where GNSS signals are interrupted fre-
Although the both solutions cannot achieve a high fix- quently. The MF GEC PPP-RTK always performs better
ing percentage under such a complex environment, the than the GF GEC PPP-RTK, and much better than the
MF GEC PPP-RTK still greatly improved the fixing per- DF GPS-only PPP-RTK and the DF GEC PPP. Therefore
centage from 40.38% to 80.77%. After the vehicle left the the proposed multi-frequency and multi-GNSS PPP-RTK
tunnel, an instantaneous ambiguity resolution can be method can greatly improve the availability of vehicle
achieved with the MF GEC PPP-RTK while it requires navigation in urban environments.
more time with dual-frequency mode.
Acknowledgements
The numerical calculations in this paper have been done on the supercom-
puting system in the Supercomputing Center of Wuhan University.
Conclusions
In this contribution, we developed a multi-frequency and Authors’ contributions
multi-GNSS PPP-RTK method aiming to achieve rapid XL and BW proposed the idea, processed the GNSS data of vehicle navigation,
and wrote this paper. XL and HL assisted in paper writing and revision. JH
positioning with the centimeter level accuracy for vehi- and XH contribute to the data acquisition and discussion on data analysis. All
cle navigation in urban environments. In the proposed authors read and approved the final manuscript.
method, with multi-frequency and multi-GNSS obser-
Funding
vations the precise atmospheric corrections are derived This work has been supported by the National Natural Science Foundation
from the server side and disseminated to the user side to of China (Grant 41974027 and Grant 41974029), the Sino-German mobility
achieve the aim. program (Grant No. M0054), and the Technology Innovation Special Project
(Major program) of Hubei Province of China (Grant No. 2019AAA043).
Beijing regional network was selected to extract the
precise atmospheric corrections. The internal accuracy Availability of data and materials
of atmospheric corrections is better than 1 cm. Vehi- The datasets used and analyzed in this study are available from the corre-
sponding author on reasonable request.
cle experiments were conducted in suburban and urban
environments to verify the performance of the multi-
Declarations
frequency and multi-GNSS PPP-RTK. In the suburbs, the
MF GEC PPP-RTK performs best. Compared to the DF Competing interests
GPS-only PPP-RTK solution, the positioning accuracy is The authors declare that they have no competing interests.
improved from the RMS of 0.134, 0.159, and 0.280 m to Author details
that of 0.019, 0.017, 0.035 m in the east, north and ver- 1
School of Geodesy and Geomatics, Wuhan University, 129 Luoyu Road,
tical components with an improvement of 87.58%, and Wuhan 430079, China. 2 Technische Universitat Berlin (TUB), Straße des 17.
Juni, 10623 Berlin, Germany.
the fixing percentage is increased from 68.60% to 93.70%.
In an urban environment where GNSS signals are inter- Received: 8 December 2021 Accepted: 2 March 2022
fered frequently, the MF GEC PPP-RTK still provides the
most precise, reliable and continuous positioning results,
because of the enhancement of the multi-frequency
signals. Compared to the DF GEC PPP-RTK solution, References
the positioning accuracy is improved from the RMS of Cabinet Office. 2020. Quasi-Zenith Satellite System performance standard (PS-
QZSS-002). https://​qzss.​go.​jp/​en/​techn​ical/​downl​oad/​pdf/​ps-​is-​qzss/​ps-​qzss-​
0.217, 0.217, and 0.346 m to the RMS of 0.040, 0.052, and 002.​pdf?t=​16180​40032​826.
0.134 m in the east, north and vertical components, with European GNSS Agency (2019). PPP-RTK market and technology report.
the fixing percentage improved from 62.66% to 83.94%.
Li et al. Satellite Navigation (2022) 3:7 Page 11 of 11

Geng, J., & Bock, Y. (2013). Triple-frequency GPS precise point positioning with rapid Zhang, B., Teunissen, P. J. G., & Odijk, D. (2011). A novel un-differenced PPP-RTK
ambiguity resolution. Journal of Geodesy, 87(5), 449–460. concept. Journal of Navigation, 64(S1), S180–S191.
Geng, J., Guo, J., Meng, X., et al. (2020). Speeding up PPP ambiguity resolution using Zhang, B., Chen, Y., & Yuan, Y. (2019). PPP-RTK based on undifferenced and uncom-
triple-frequency GPS/BeiDou/Galileo/QZSS data. Journal of Geodesy, 94, 6. bined observations: Theoretical and practical aspects. Journal of Geodesy, 93(7),
https://​doi.​org/​10.​1007/​s00190-​019-​01330-1 1011–1024.
Gu, S., Lou, Y., Shi, C., & Liu, J. (2015). BeiDou phase bias estimation and its applica- Zhang, B., Ou, J., Yuan, Y., et al. (2012). Extraction of line-of-sight ionospheric observa-
tion in precise point positioning with triple-frequency observable. Journal of bles from GPS data using precise point positioning. Science China Earth
Geodesy, 89(10), 979–992. Sciences, 55, 1919–1928. https://​doi.​org/​10.​1007/​s11430-​012-​4454-8
Guo, F., Zhang, X., & Wang, J. (2015). Timing group delay and differential code bias Zhang, X., & Li, P. (2013). Assessment of correct fixing rate for precise point position-
corrections for BeiDou positioning. Journal of Geodesy, 89(5), 427–445. ing ambiguity resolution on a global scale. Journal of Geodesy, 87, 579–589.
Jokinen, A., Feng, S., Schuster, W., Ochieng, W., Hide, C., Moore, T., & Hill, C. (2013). Zumberge, J. F., Heflin, M. B., Jefferson, D. C., Watkins, M. M., & Webb, F. H. (1997).
GLONASS aided GPS ambiguity fixed precise point positioning. Journal of Precise point positioning for the efficient and robust analysis of GPS data from
Navigation, 66(3), 399–416. large networks. Journal of Geophysical Research, 102(B3), 5005–5017. https://​
Kouba, J., & Heroux, P. (2001). Precise point positioning using IGS orbit and clock doi.​org/​10.​1029/​96JB0​3860
products. GPS Solut, 5(2), 12–28.
Kouba J. 2015. A guide to using international GNSS service (IGS) products. http://​kb.​
igs.​org/​hc/​en-​us/​artic​les/​20127​1873-A-​Guide-​to-​Using-​the-​IGS-​Produ​cts. Publisher’s Note
Khodabandeh, A., & Teunissen, P. J. G. (2016). PPP-RTK and inter-system biases: The Springer Nature remains neutral with regard to jurisdictional claims in pub-
ISB look-up table as a means to support multi-system PPP-RTK. Journal of lished maps and institutional affiliations.
Geodesy, 90(9), 837–851. https://​doi.​org/​10.​1007/​s00190-​016-​0914-9
Li, X., Zhang, X., & Ge, M. (2011). Regional reference network augmented precise
point positioning for instantaneous ambiguity resolution. Journal of Geodesy,
85(3), 151–158. https://​doi.​org/​10.​1007/​s00190-​010-​0424-0
Li, X., Ge, M., Zhang, H., & Wickert, J. (2013). A method for improving uncalibrated
phase delay estimation and ambiguity-fixing in real-time precise point
positioning. Journal of Geodesy, 87, 405–416. https://​doi.​org/​10.​1007/​
s00190-​013-​0611-x
Li, X., Ge, M., Douša, J., & Wickert, J. (2014). Real-time precise point positioning
regional augmentation for large GPS reference networks. GPS Solutions, 18(1),
61–71.
Li, X., Li, X., Yuan, Y., Zhang, K., Zhang, X., & Wickert, J. (2018). Multi-GNSS phase delay
estimation and PPP ambiguity resolution: GPS, BDS, GLONASS, Galileo. Journal
of Geodesy, 92, 579–608. https://​doi.​org/​10.​1007/​s00190-​017-​1081-3
Li, X., Li, X., Liu, G., et al. (2019). Triple-frequency PPP ambiguity resolution with
multi-constellation GNSS: BDS and Galileo. Journal of Geodesy, 93, 1105–1122.
https://​doi.​org/​10.​1007/​s00190-​019-​01229-x
Li, X., Han, X., Li, X., et al. (2021). GREAT-UPD: An open-source software for uncali-
brated phase delay estimation based on multi-GNSS and multi-frequency
observations. GPS Solut, 25, 66. https://​doi.​org/​10.​1007/​s10291-​020-​01070-2
Liu, Y., Song, W., Lou, Y., Ye, S., & Zhang, R. (2017). GLONASS phase bias estimation
and its PPP ambiguity resolution using homogeneous receivers. GPS Solutions,
21(2), 427–437.
Montenbruck, O., Hauschild, A., Steigenberger, P., & Langley, R. B. (2010). Three’s the
challenge: A close look at GPS SVN62 triple-frequency signal combinations
finds carrier-phase variations on the new L5. GPS World, 21(8), 8–19.
Montenbruck O, Hauschild A, Steigenberger P. (2014). Differential code bias estima-
tion using multi‐GNSS observations and global ionosphere maps. Navigation:
Journal of The Institute of Navigation 61(3), 191–201.
de Oliveira, P. S., Morel, L., Fund, F., Legros, R., Monico, J. F. G., Durand, S., & Durand,
F. (2017). Modeling tropospheric wet delays with dense and sparse network
configurations for PPP-RTK. GPS Solutions, 21(1), 237–250.
Pan, L., Zhang, X., Li, X., Liu, J., & Li, X. (2017). Characteristics of inter-frequency clock
bias for Block IIF satellites and its effect on triple-frequency GPS precise
point positioning. GPS Solutions, 21(2), 811–822. https://​doi.​org/​10.​1007/​
s10291-​016-​0571-8
Pan, L., Li, X., Zhang, X., Li, X., Lu, C., Zhao, Q., & Liu, J. (2017). Considering inter-fre-
quency clock bias for BDS triple-frequency precise point positioning. Remote
Sensing, 9(7), 734. https://​doi.​org/​10.​3390/​rs907​0734
Pan, L., Zhang, X., Guo, F., & Liu, J. (2019). GPS inter-frequency clock bias estima-
tion for both uncombined and ionospheric-free combined triple-frequency
precise point positioning. Journal of Geodesy, 93(4), 473–487.
Ren, X., Zhang, J., Chen, J., Zhang, X.(*) (2021). Ionospheric Modeling by using Multi-
GNSS and upcoming LEO Constellations: Two Methods and Comparison. In
IEEE Transactions on Geoscience and Remote Sensing, 1–15
Teunissen, P., Odijk, D., & Zhang, B. (2010). PPP-RTK: Results of CORS network-based
PPP with integer ambiguity resolution. J Aeronaut Astronaut Aviat Ser A, 42(4),
223–230.
Wubbena G, Schmitz M, Bagg A (2005) PPP-RTK: Precise point positioning using
state-space representation in RTK networks. In: Proceedings of ION GNSS, pp
13–16.