Modeling and Prediction of Inter-System Bias for GPS/BDS-2/BDS-3 Combined Precision Point Positioning

The combination of Precision Point Positioning(PPP)with Multi-Global Navigation Satellite System(MultiGNSS),called MGPPP,can improve the positioning precision and shorten the convergence time more effectively than the combination of PPP with only the BeiDou Navigation Satellite System(BDS).However,the Inter-System Bias(ISB)measurement of Multi-GNSS,including the time system offset,the coordinate system difference,and the inter-system hardware delay bias,must be considered for Multi-GNSS data fusion processing.The detected ISB can be well modeled and predicted by using a quadratic model(QM),an autoregressive integrated moving average model(ARIMA),as well as the sliding window strategy(SW).In this study,the experimental results indicate that there is no apparent difference in the ISB between BDS-2 and BDS-3 observations if B1I/B3I signals are used.However,an obvious difference in ISB can be found between BDS-2 and BDS-3 observations if B1I/B3I and B1C/B2a signals are used.Meanwhile,the precision of the Predicted ISB(PISB)on the next day of all stations is about 0.1−0.6 ns.Besides,to effectively utilize the PISB,a new strategy for predicting the PISB for MGPPP is proposed.In the proposed strategy,the PISB is used by adding two virtual observation equations,and an adaptive factor is adopted to balance the contribution of the Observed ISB(OISB)and the PISB to the final estimations of ISB.To validate the effectiveness of the proposed method,some experimental schemes are designed and tested under different satellite availability conditions.The results indicate that in open sky environment,the selective utilization of the PISB achieves almost the same positioning precision of MGPPP as the direct utilization of the PISB,but the convergence time of MGPPP is reduced by 7.1%at most in the north(N),east(E),and up(U)components.In the blocked sky environment,the selective utilization of the PISB contributes to more significant improvement of the positioning precision and convergence time than that in the open sky environment.Compared with the direct utilization of the PISB,the selective utilization of the PISB improves the positioning precision and convergence time by 6.7%and 12.7%at most in the N,E,and U components,respectively.

The mathematical weighting of GNSS observations based on different types of receivers/antennas and environmental conditions

Stochastic models play an important role in achieving high accuracy in positioning,the ideal estimator in the least-squares(LS)can be obtained only by using the suitable stochastic model.This study investigates the role of variance component estimation(VCE)in the LS method for Precise Point Positioning(PPP).This estimation is performed by considering the ionospheric-free(IF)functional model for code and the phase observation of Global Positioning System(GPS).The strategy for estimating the accuracy of these observations was evaluated to check the effect of the stochastic model in four modes:a)antenna type,b)receiver type,c)the tropospheric effect,and d)the ionosphere effect.The results show that using empirical variance for code and phase observations in some cases caused erroneous estimation of unknown components in the PPP model.This is because a constant empirical variance may not be suitable for various receivers and antennas under different conditions.Coordinates were compared in two cases using the stochastic model of nominal weight and weight estimated by LS-VCE.The position error difference for the east-west,north-south,and height components was 1.5 cm,4 mm,and 1.8 cm,respectively.Therefore,weight estimation with LS-VCE can provide more appropriate results.Eventually,the convergence time based on four elevation-dependent models was evaluated using nominal weight and LS-VCE weight.According to the results,the LS-VCE has a higher convergence rate than the nominal weight.The weight estimation using LS-VCE improves the convergence time in four elevation-dependent models by 11,13,12,and 9 min,respectively.

Obtaining accurate measurements of the sea surface height from a GPS buoy

A dedicated GPS buoy is designed for calibration and validation(Cal/Val)of satellite altimeters since 2014.In order to evaluate the accuracy of the sea surface height(SSH)measured by the GPS buoy,twelve campaigns have been done within China sea area between 2014 and 2021.In six of these campaigns,two static Global Navigation Satellite System stations were installed at distances of<1 km and 19 km from the buoy to assess how the baseline length influenced the derived SSH from the buoy solutions.The GPS buoy data was processed using the GAMIT/GLOBK software+TRACK module and CSRS-PPP tool to achieve the SSH.The SSH was compared with conventionally tide gauge(TG)data to evaluate the accuracy of the buoy with the standard deviation of the height element.The results showed that the difference in the standard deviation of the SSH from the buoy and the TG was less than 16 mm.The SSHs processed with different ephemeris(Ultra-Rapid,Rapid,Final)were not significantly different.When the baseline length was 19 km,the SSH solution of the GPS buoy performed well,with standard bias of less than 26 mm between the heights measured by the buoy and TG,meaning that the buoy could be used for Cal/Val of altimeters.The bias between the Canadian Spatial Reference System-precise point positioning tool and the TRACK varied a lot,and some of them were over 130 mm.This deemed too high to be useful for Cal/Val of satellite altimeters.Moreover,the GPS buoy solutions processed by GAMIT/GLOBK software+TRACK module were used for in-orbit Cal/Val of HY-2B/C satellites in ten campaigns.The SSH and significant wave height of the altimeters showed good agreements with the GPS buoy solutions.

An ERA5 tropospheric parameters-augmented approach for improving GNSS precise point positioning

Precise Point Positioning(PPP) technology has developed into a potent instrument for geodetic positioning, ionospheric modeling, tropospheric atmospheric parameter detection, and seismic monitoring.As atmospheric reanalysis data products’ accuracy and spatiotemporal resolution have improved recently, it has become important to apply these products to obtain high-accuracy tropospheric delay parameters, like zenith tropospheric delay(ZTD) and tropospheric horizontal gradient. These tropospheric delay parameters can be applied to PPP to reduce the convergence time and to increase the accuracy in the vertical direction of the position. The European Centre for Medium-Range Weather Forecasts Reanalysis 5(ERA5) atmospheric reanalysis data is the latest product with a high spatiotemporal resolution released by the European Center for Medium-Range Weather Forecasts(ECMWF). Only a few researches have evaluated the application of ERA5 data to Global Navigation Satellite System(GNSS)PPP. Therefore, this study compared and validated the ZTD products derived from ERA5 data using ZTD values provided by 290 global International GNSS Service(IGS) stations for 2016-2017. The results indicated a stable performance for ZTD, with annual average bias and RMS values of 0.23 cm and 1.09 cm,respectively. Further, GNSS observations for one week in each of the four seasons(spring: DOY 92-98;summer: DOY 199-205;autumn: DOY 275-281;and winter: DOY 22-28) from 34 multi-GNSS experiments(MGEX) stations distributed globally in 2016 were considered to evaluate the performance of ERA5-derived tropospheric delay products in GNSS PPP. The performance of ERA5-enhanced PPP was compared with that of the two standard GNSS PPP schemes(without estimated tropospheric horizontal gradient and with estimated tropospheric horizontal gradient). The results demonstrated that ERA5-enhanced GNSS PPP showed no significant improvement in the convergence times in both the Eastern(E) and Northern(N) directions, while the average convergence time over four weeks in the vertical(U)direction improved by 53.3% and 52.7%, respectively(in the case of pngm station). The average convergence times for each week in the U direction of the northern and southern hemisphere stations indicated a decrease of 16.3%, 12.6%, 9.6%, and 9.1%, and 16.9%, 9.6%, 8.9%, and 14.5%, respectively.Regarding positioning accuracy, ERA5-enhanced PPP showed an improvement of 13.3% and 16.2% over the two standard PPP schemes in the U direction, respectively. No significant improvement in the positioning performance was observed in both the E and N directions. Thus, this study demonstrated the potential application of the ERA5 tropospheric parameters-augmented approach to Beidou navigation and positioning.

Status of UnDifferenced and Uncombined GNSS Data Processing Activities in China

With the continued development of multiple Global Navigation Satellite Systems(GNSS)and the emergence of various frequencies,UnDifferenced and UnCombined(UDUC)data processing has become an increasingly attractive option.In this contribution,we provide an overview of the current status of UDUC GNSS data processing activities in China.These activities encompass the formulation of Precise Point Positioning(PPP)models and PPP-Real-Time Kinematic(PPP-RTK)models for processing single-station and multi-station GNSS data,respectively.Regarding single-station data processing,we discuss the advancements in PPP models,particularly the extension from a single system to multiple systems,and from dual frequencies to single and multiple frequencies.Additionally,we introduce the modified PPP model,which accounts for the time variation of receiver code biases,a departure from the conventional PPP model that typically assumes these biases to be time-constant.In the realm of multi-station PPP-RTK data processing,we introduce the ionosphere-weighted PPP-RTK model,which enhances the model strength by considering the spatial correlation of ionospheric delays.We also review the phase-only PPP-RTK model,designed to mitigate the impact of unmodelled code-related errors.Furthermore,we explore GLONASS PPP-RTK,achieved through the application of the integer-estimable model.For large-scale network data processing,we introduce the all-in-view PPP-RTK model,which alleviates the strict common-view requirement at all receivers.Moreover,we present the decentralized PPP-RTK data processing strategy,designed to improve computational efficiency.Overall,this work highlights the various advancements in UDUC GNSS data processing,providing insights into the state-of-the-art techniques employed in China to achieve precise GNSS applications.

Deformation caused by the 2011 eastern Japan great earthquake monitored using the GPS single-epoch precise point positioning technique

Crustal deformation can provide constraints for studying earthquake rupture and shock wave transmission for the Mw9.0 eastern Japan great earthquake. Using the single- epoch precise point positioning （PPP） method and the appropriate positioning flow, we process GPS data from six IGS （International GNSS Service） sites （e.g., MIZU, TSK2, USUD, MTKA, AIRA and KSMV） located in Japan and obtain the positioning results with centimeter scale precision. The displacement time series of the six sites are analyzed using the least squares spectral analysis method to estimate deformations caused by the Mw9.0 mainshock and the Mw7.9 aftershock, and the cumulative displacements after 1 day. Mainshock displacements at station MIZU, the nearest site to the mainshock in the North （N）, East （E）, and Up （U） directions, are -1.202 m, 2.180 m and -0.104 m, respectively, and the cumulative deformations after 1 day are -1.117 m, 2.071 m and -0.072 m, respectively. The displacements at station KSMV, the nearest site to the Mw7.9 aftershock in the N, E and U directions, are -0.032 m, 0.742 m and -0.345 m, respectively. The other sites obviously experienced eastern movements and subsidence. The deformation vectors indicate that the horizontal displacements caused by the earthquake point to the epicenter and rupture. Elastic bounds evidently took place at all sites. The results indicate that the crustal movements and earthquake were part of a megathrust caused by the Pacific Plate sinking under the North American Plate to the northeast of Japan island arc.

Modified algorithm of combined GPS/GLONASS precise point positioning for applications in open-pit mines

A modified algorithm of combined GPS/GLONASS precise point positioning （GG-PPP） was developed by decreasing the number of unknowns to be estimated so that accurate position solutions can be achieved in the case of less number of visible satellites. The system time difference between GPS and GLONASS （STDGG） and zenith tropospheric delay （ZTD） values were firstly estimated in an open sky condition using the traditional GG-PPP algorithm. Then, they were used as a priori known values in the modified algorithm instead of estimating them as unknowns. The proposed algorithm was tested using observations collected at BJFS station in a simulated open-pit mine environment. The results show that the position filter converges much faster to a stable value in all three coordinate components using the modified algorithm than using the traditional algorithm. The modified algorithm achieves higher positioning accuracy as well. The accuracy improvement in the horizontal direction and vertical direction reaches 69% and 95% at a satellite elevation mask angle of 50°, respectively.

Signal quality analysis and quality check of BDS3 Precise Point Positioning in the Arctic Ocean

This study analyzes the signal quality and the accuracy of BeiDou 3 rd generation Satellite Navigation System(BDS3) Precise Point Positioning(PPP) in the Arctic Ocean. Assessment of signal quality of BDS3 includes signal to noise ratio(SNR), multipath(MP), dilution of precision(DOP), and code-minus-carrier combination(CC). The results show that, 5 to 13 satellites are visible at any time in the Arctic Ocean area as of September 2018, which are sufficient for positioning. In the mid-latitude oceanic region and in the Arctic Ocean, the SNR is 25–52 dB Hz and the MP ranges from-2 m to 2 m. As the latitude increases, the DOP values show large variation, which may be related to the distribution of BDS satellites. The CC values of signals B1 I and BIC range from-5 m to 5 m in the mid-latitude sea area and the Arctic Ocean, which means the effect of pseudorange noise is small. Moreover, as to obtain the external precise reference value for GNSS positioning in the Arctic Ocean region is difficult, it is hard to evaluate the accuracy of positioning results. An improved isotropy-based protection level method based on Receiver Autonomous Integrity Monitoring is proposed in the paper, which adopts median filter to smooth the gross errors to assess the precision and reliability of PPP in the Arctic Ocean. At first, the improved algorithm is verified with the data from the International GNSS Service Station Tixi. Then the accuracy of BDS3 PPP in the Arctic Ocean is calculated based on the improved algorithm. Which shows that the kinematic accuracy of PPP can reach the decimeter level in both the horizontal and vertical directions, and it meets the precision requirements of maritime navigation.

Estimation of annual variation of water vapor in the Arctic Ocean between 80°–87°N using shipborne GPS data based on kinematic precise point positioning

The measurement of atmospheric water vapor （WV） content and variability is important for meteorological and climatological research. A technique for the remote sensing of atmospheric WV content using ground-based Global Positioning System （GPS） has become available, which can routinely achieve accuracies for integrated WV content of 1-2 kg/m2. Some experimental work has shown that the accuracy of WV measurements from a moving platform is comparable to that of （static） land-based receivers. Extending this technique into the marine environment on a moving platform would be greatly beneficial for many aspects of meteorological research, such as the calibration of satellite data, investigation of the air-sea interface, as well as forecasting and climatological studies. In this study, kinematic precise point positioning has been developed to investigate WV in the Arctic Ocean （80°-87°N） and annual variations are obtained for 2008 and 2012 that are identical to those related to the enhanced greenhouse effect.

Velocity of Surface Ice Flow on Amery Ice Shelf Determined with PPP

The main activities in the joint expedition between CHINARE and ANARE on Amery ice shelf are introduced. Five day continuous GPS observation data collected on the site which locates at the frontal part of Amery ice shelf was processed with precise point positioning (PPP) technology based on precise products from IGS. Velocity of the surface ice flow on Amery can be derived from the PPP solution. Preliminary result shows that the surface ice flow velocity of the site is 2.25 meters per day, the motion direction is northeastward. Semidiurnal oceanic tide and diurnal oceanic tide signal of that site can be recovered from the height variation series of PPP solution. These above solutions can be used to the consequent mass balance calculation.

Land deformation monitoring in mining area with PPP-AR

The mining area deformation monitoring theory and method using precise point positioning （PPP） ambi- guity resolution （AR） were studied, and an ambiguity fixing model with satellite and receiver combina- tion phase delay （CPD） was proposed for zero-differenced PPP ambiguity fixing and its corresponding formula derivation was given. The data processing results for I h at six IGS stations in China show that 93% of ambiguities can be fixed within 10 min and all ambiguities can be fixed within 15 min. After ambi- guity fixing, the positioning accuracy is improved by more than 85% in the E and N directions, with abso- lute positioning accuracy reaching millimeter level, and it was improved by 70% in the U direction, reaching centimeter level; the proposed zero-differenced ambiguity fixing model can effectively improve the convergence rate and positioning accuracy in PPP. Data monitoring continuously conducted for half a year at four COPS stations of Shanxi China Coal Pingshuo Group validated the feasibility of using PPP in mining area deformation monitoring.

Comparison of availability and reliability among different combined-GNSS/RNSS precise point positioning

With emergence of the BeiDou Navigation Satellite System(BDS), the Galileo Satellite Navigation System(Galileo), the Quasi-Zenith Satellite System(QZSS)and the restoration of the Global Navigation Satellite System(GLONASS), the single Global Positioning System(GPS) has been gradually expanded into multiple global and regional navigation satellite systems(multi-GNSS/RNSS). In view of differences in these 5 systems, a consolidated multi-GNSS/RNSS precise point positioning(PPP) observation model is deduced in this contribution. In addition, the performance evaluation of PPP for multi-GNSS/RNSS is conducted using a large number of the multi-GNSS experiment(MGEX) station datasets. Experimental results show that multi-GNSS/RNSS can guarantee plenty of visible satellites effectively. Compared with single-system GPS, PDOP, HDOP, and VDOP values of the multi-GNSS/RNSS are improved by 46.8%, 46.5% and 46.3%, respectively. As for convergence time, the static and kinematic PPP of multi-GNSS/RNSS are superior to that of the single-system GPS, whose reliability, availability, and stability drop sharply with the increasing elevation cutoff. At satellite elevation cutoff of 40 °, the single-system GPS fails to carry out continuous positioning because of the insufficient visible satellites, while the multi-GNSS/RNSS PPP can still get positioning solutions with relatively high accuracy, especially in the horizontal direction.

Impact of Tropospheric Delay Gradients on Total Tropospheric Delay and Precise Point Positioning

GPS signals are electromagnetic waves that are affected by the Earth’s atmosphere. The Earth’s atmosphere can be categorized, according to its effect on GPS signals, into the ionosphere (ionospheric delay) and neutral atmosphere (tropospheric delay). The first-order ionospheric delay can be eliminated by linear combination of GPS observables on different frequencies. However, tropospheric delay cannot be eliminated because it is frequency-independent. The total tropospheric delay can be divided into three components. The first is the dry component, the second part is the wet component, and the third part is the horizontal gradients which account for the azimuthal dependence of tropospheric delay. In this paper, the effect of modeling tropospheric gradients on the estimation of the total tropospheric delay and station position is investigated. Long session, one month during January 2015, of GPS data is collected from ten randomly selected globally distributed IGS stations. Two cases are studied: the first case, the coordinates of stations are kept fixed to their actual values and the tropospheric delay is estimated twice, with and without tropospheric gradients. In the second case, the station position is estimated along with the total tropospheric delay with and without tropospheric gradients. It is shown that the average bias of the estimated total tropospheric delay when neglecting tropospheric gradients ranges from ?1.72 mm to 2.14 mm while the average bias when estimating gradients are ?0.898 mm to 1.92 mm which means that the bias is reduced by about 30%. In addition, the average standard deviation of the bias is 4.26 mm compared with 4.52 mm which means that the standard deviation is improved by about 6%.

Stochastic Analysis of Low-Cost Single-Frequency GPS Receivers

Typically, dual-frequency geodetic grade GNSS receivers are utilized for positioning applications that require high accuracy. Single-frequency high grade receivers can be used to minimize the expenses of such dual-frequency receivers. However, user has to consider the resultant positioning accuracy. Since the evolution of low-cost single-frequency (LCSF) receivers is typically cheaper than single-frequency high grade receivers, it is possible to obtain comparable positioning accuracy if the corresponding observables are accurately modelled. In this paper, two LCSF GPS receivers are used to form short baseline. Raw GPS measurements are recorded for several consecutive days. The collected data are used to develop the stochastic model of GPS observables from such receivers. Different functions are tested to determine the best fitting model which is found to be 3 parameters exponential decay function. The new developed model is used to process different data sets and the results are compared against the traditional model. Both results from the newly developed and the traditional models are compared with the reference solution obtained from dual-frequency receiver. It is shown that the newly developed model improves the root-mean-square of the estimated horizontal coordinates by about 10% and improves the root-mean-square of the up component by about 39%.

Equivalence of Network-Solution and PPP-Solution

The technique of precise point positioning (PPP) is gradually becoming a popular method in GPS data-processing. In GPS observation equation, the unknown parameters can be separated into two parts: global parameters and local parameters. The global parameters include orbit, satellite clock and geodynamic parameters. The local parameters are site-occupation-specific, such as position, tropospheric delay, etc. The formulas of local parameters are firstly derived under the network-solution and the PPP-solution conditions respectively. If the weight matrix of global parameters in PPP-solution is small enough, the cofactor matrices of local parameters are the same as that in network-solution. Then, 16 daily solutions are obtained in both PPP mode and network mode. Three sites are selected to compare the solutions. The experimental results demonstrated that the difference between two solutions in coordinates and tropospheric delays are only few millimeters. This level of difference can be neglected so that the solutions from both PPP mode and network mode can be taken as the same in the actual application.

Application of precise point positioning technology in airborne gravity measurement

The precise point positioning （PPP） technology is applied to an airborne gravity survey. By analyzing the advantages and disadvantages of several velocity and acceleration measurement methods and in combination with an actual marine gravity survey, the position difference method is confirmed to be a useful survey method for velocity and acceleration. Finally, the practicability of using PPP in airborne marine gravity survey is verified by measured data.

Performance analysis of real-time and post-mission kinematic precise point positioning in marine environments

This article focuses on the performance analysis of both real-time and post-mission kinematic precise point positioning(PPP)in challenging marine environments.For this purpose,a real dynamic experiment lasting 6 h was carried out on a lake dam in?orum City of Turkey.While the kinematic test was continuing,the real-time PPP coordinates were obtained for each measurement epoch with a commercial real-time PPP(RT-PPP)service,namely the Trimble Center Point RTX.Then the post-mission PPP(PM-PPP)coordinates were calculated by using Multi-GNSS data and the Multi-GNSS Experiment(MGEX)precise products.The kinematic RT-PPP and PM-PPP results showed that the PPP coordinates were consistent with the relative solution at centimetre and decimetre level in horizontal and height components,respectively.This study implies that PPP technique is a powerful tool for highly accurate positioning in both real-time and post-mission modes,even for dynamic applications in harsh environments.

A rapid stability assessment of China's IGS sites after the Ms7.0 Lushan earthquake

A rapid and accurate assessment of the stability of surveying and mapping reference points is important for post - disaster rescue, disaster relief and reconstruction activities. Using Precise Point Positioning （PPP） technology, a rapid assessment of the stability of the IGS sites in China was performed after the Ms7.0 Lushan earthquake using rapid precise ephemeris and rapid precise satellite clock products. The results show that the earthquake had a very small impact and did not cause significant permanent deformation at the IGS sites. Most of the sites were unaffected and remained stable after the earthquake.

基于伪距/相位OSB的多频多模PPP-AR及其性能评估

在进行多频多模全球导航卫星系统(global navigation satellite system,GNSS)精密单点定位(precise point positioning,PPP)时,以相对形式表达的伪距差分码偏差(differential code bias,DCB)和相位小数周偏差(fractional cycle bias,FCB)种类繁多且修正方法较为复杂。基于此,首先给出了相对形式的伪距/相位偏差与原始观测值上的绝对偏差(observable-specific signal bias,OSB)的转换方法,以及利用该偏差进行非差模糊度固定的具体流程;然后通过实测数据分析不同偏差形式下的PPP模糊度解算(PPP-ambiguity resolution,PPP-AR)定位性能。结果表明,无论是多系统PPP还是多频PPP,其基于OSB的PPP-AR定位精度、收敛时间及固定率均与相对形式DCB/FCB的PPP-AR定位结果相当。其中,三频PPP-AR的动态和静态收敛时间均优于双频,多系统PPP-AR定位的初始化时间和定位精度也均优于单系统。伪距/相位OSB可直接用于修正对应的观测值,减少偏差处理中的组合与转换,并简化GNSS PPP数据处理流程,提高了频率选择的灵活性,使用户端算法更简单方便。

GNSS rapid precise point positioning enhanced by low Earth orbit satellites

The Low Earth Orbit(LEO)satellites can be used to efectively speed up Precise Point Positioning(PPP)convergence.In this study,180 LEO satellites with a global distribution are simulated to evaluate their contribution to the PPP convergence.LEO satellites can give more redundant observations and improve satellite geometric distributions,particularly for a single Global Navigation Satellite System(GNSS).The convergence speed of the PPP foat solution using the Global Positioning System(GPS,G)or BeiDou Navigation Satellite System(BDS,C)single system as well as the G/C/Galileo navigation satellite system(Galileo,E)/GLObal NAvigation Satellite System(GLONASS,R)combined system with LEO satellites added is improved by 90.0%,91.0%,and 90.7%,respectively,with respect to the system without LEO satellites added.We introduced LEO observations to assist GNSS in PPP-AR(Ambiguity Resolution)and PPP-RTK(Real Time Kinematic).The success fx rate of a single system is signifcantly improved,and the Time-To-First-Fix(TTFF)of G and G/C/E is reduced by 86.4%and 82.8%,respectively,for the PPP-AR solution.We analyzed the positioning performance of LEO satellite assisted G/C/E PPP-RTK in the reference networks of diferent scales,namely diferent atmospheric delay interpolation accuracies.The success fx rate of the G/C/E combined system is improved from 86.8 to 94.9%,and the TTFF is reduced by 36.8%,with the addition of LEO satellites in the 57 km reference network.In the 110 km reference network,the success fx rate of the G/C/E combined system is improved from 64.0 to 88.6%,and the TTFF is reduced by 32.1%.GNSS PPP-RTK with adding the LEO satellites in the reference networks of diferent scales shows obvious improvement because the atmospheric correlation decreases with increasing distance from the reference networks.