A device and method for differential calibration of a time transfer link based on GNSS analog signals

Abstract

A kind of time transfer link differential calibration device and method based on GNSS analog signal, for the defects that traditional real signal differential calibration is interfered by atmosphere, multipath, long observation period, existing absolute calibration experiment complexity and equipment threshold, build the common clock differential system containing time frequency distribution module, GNSS simulator, microwave dark box, reference / dual time transfer link to be measured, counter and vector network analyzer;Simulator closes all propagation error, dark box ultra-short baseline synchronous reception simulation signal, after collecting 1~2 days of observation data, solve common clock difference, solve the hardware delay of to-be-calibrated time-frequency transfer receiver using differential formula and evaluate uncertainty in layers, which can be cross-verified with differential calibration and two types of absolute calibration.The experimental environment of the present application is controllable and reproducible, eliminates atmospheric interference, does not need to calibrate the time delay of simulator, the uncertainty of Beidou multi-frequency point calibration synthesis is better than 0.94 ns, and it is suitable for small and medium-sized metrology laboratories to carry out Beidou, GPS time-frequency transfer receiver batch hardware delay calibration.

Description

Technical Field

[0001] This invention belongs to the field of time and frequency metrology and satellite navigation time and frequency transmission calibration technology. Specifically, it relates to a device and measurement method that uses a GNSS signal simulator and a microwave anechoic chamber to construct a controllable laboratory environment, combined with a common clock difference (CCD) experiment to achieve precise differential calibration of hardware delay in the time transmission link. It is applicable to full-frequency hardware delay calibration of time and frequency transmission receivers for multi-mode GNSS (Global Navigation Satellite System) such as BeiDou, GPS, and GLONASS. Background Technology

[0002] Currently, the mainstream time and frequency transfer methods internationally include three types: satellite time and frequency transfer, represented by Global Navigation Satellite System (GNSS) time and frequency transfer and Two-Way Satellite Time and Frequency Transfer (TWSTFT), and fiber optic time and frequency transfer. Due to its advantages such as wide coverage, fast real-time information transmission, lower equipment cost, and higher transmission and synchronization accuracy, GNSS time and frequency transfer has gradually become the most widely used long-distance time and frequency transfer method, and is the best choice for long-distance time tracing.

[0003] A GNSS time-frequency transfer receiver typically consists of three hardware components: a GNSS time-frequency transfer receiver antenna, an antenna feeder, and the GNSS time-frequency transfer receiver main unit. It can be used to establish a comparison link with a local time-frequency reference source, obtaining the time deviation between the local time standard and the GNSS system based on pseudorange or carrier phase observations. Different ground observation stations can also perform long-baseline time-frequency comparisons using methods such as GNSS Common View (CV), All View (AV), and Precise Point Positioning (PPP), thereby achieving time-frequency value transfer between the two stations. In GNSS time-frequency transfer and the extended applications of the above methods, the GNSS time-frequency transfer receiver and related equipment extensively utilize electrical components, resulting in significant hardware delays. Therefore, the time-frequency comparison values ​​calculated based on pseudorange or carrier phase observations cannot directly and accurately reproduce the reference GNSS time standard or reflect the exact time difference between external time reference sources. Precise measurement of hardware delay is necessary to compensate for system errors caused by hardware delay in the GNSS time-frequency transfer receiver during time transmission; if users cannot perform precise calibration of the hardware delay in the GNSS time-frequency transfer receiver, they cannot achieve accurate time transmission and synchronization.

[0004] Currently, there are three methods for GNSS time transfer link calibration: differential calibration based on real GNSS signals, step-by-step absolute calibration based on GNSS simulator signals, and overall absolute calibration. The differential calibration method is researched and led by the International Bureau of Weights and Mesures (BIPM). It utilizes a calibrated reference time transfer link and the time transfer link to be calibrated to perform a common clock difference (CCD) experiment. Long-term observation data from both are processed and calculated, and the sum of the hardware delay of the time-frequency transfer receiver host and the GNSS receiving antenna delay is obtained after deducting the individual delays of other parts of the time transfer link. However, current international differential calibration methods rely solely on real signals and require continuous recording of observation data for 4-7 days. Because real satellite observations and environmental conditions change over time, they lack the ability to be repeatedly verified under controlled laboratory conditions. In addition, the external environment is affected by the troposphere, ionosphere, and multipath effects. The zenith troposphere and ionosphere can affect the propagation distance of single-frequency signals by up to 10m and 2.5m, respectively. Although the ultra-short baseline experimental setup can reduce the influence of the troposphere and ionosphere, residual quantities and multipath effect errors still exist. The references indicate that the multipath effect contributes 0.3 ns to the uncertainty of a single time-frequency transfer receiver. Regarding absolute calibration, the step-by-step absolute calibration method was initially only studied and participated in by a few institutions. Now, several major international time laboratories and research centers have also conducted step-by-step absolute calibration experiments on time transfer links. The overall absolute calibration method is currently only studied and participated in by a few domestic laboratories. Although absolute calibration has the advantage of lower uncertainty, it requires additional measurements of simulator delay and RF transmission path delay, making it more complex and difficult to implement than differential calibration, and therefore unsuitable for small and medium-sized laboratories to carry out related calibration activities. Summary of the Invention

[0005] Traditional differential calibration methods suffer from signal interference from the ionosphere, troposphere, and multipath effects during propagation in the external environment. Furthermore, the quality of the actual signal is affected by satellite observations and the time-varying nature of satellite clock bias, lacking reproducible and repeatable verification conditions. This invention proposes a time-transfer link differential calibration device and method based on a GNSS simulator. It uses a standard GNSS simulator signal and a microwave anechoic chamber to replace the GNSS signal in the real environment, eliminating the need for 4-7 days of observation time. This provides reproducible and repeatable verification conditions in the laboratory, reducing error sources and improving the uncertainty level of the differential calibration method. Compared to absolute calibration methods, this invention leverages the advantages of differential calibration to eliminate common time delay terms, eliminating the need to measure the delay of each part of the signal transmission path individually, simplifying experimental steps and reducing experimental complexity.

[0006] The objective of this invention is achieved through the following technical solution:

[0007] A time-transfer link differential calibration device based on GNSS analog signals includes a time-frequency source allocation module, a GNSS signal simulator, a microwave anechoic chamber, a reference time-transfer link, a time-transfer link to be calibrated, a high-precision time interval counter, and a vector network analyzer.

[0008] The time-frequency source allocation module outputs a unified 10MHz frequency signal and a 1PPS second pulse synchronization signal, which are respectively connected to the GNSS signal simulator, the reference time-frequency transfer receiver, the time-frequency transfer receiver to be calibrated, and the high-precision time interval counter, so that all devices share the same time-frequency reference.

[0009] The GNSS signal simulator's radio frequency output terminal is connected to the transmitting horn antenna inside the microwave anechoic chamber. The simulator can be configured with a static carrier scene, and all error models of the ionosphere, troposphere, multipath, satellite clock error, and time-frequency transfer receiver clock error can be turned off, and the pseudorange between the satellite and the carrier can be adjusted to zero.

[0010] The microwave anechoic chamber is equipped with a transmitting horn antenna, two identical reference time-frequency transfer receiver GNSS antennas deployed with an ultra-short baseline, and a time-frequency transfer receiver GNSS antenna to be calibrated. The reference time-frequency transfer receiver GNSS antenna and the time-frequency transfer receiver GNSS antenna to be calibrated are respectively connected to the RF input ports of the reference time transfer link and the time transfer link to be calibrated through independent antenna feed lines. The inner wall of the microwave anechoic chamber is lined with absorbing material to shield external electromagnetic interference and eliminate indoor RF multipath reflections under sealed conditions.

[0011] The reference time transfer link includes a reference time and frequency transfer receiver GNSS antenna, a reference antenna feeder, and a reference time and frequency transfer receiver main unit, and stores the known hardware delay of the reference time and frequency transfer receiver. ;

[0012] The time transfer link to be calibrated includes a GNSS antenna for a time-frequency transfer receiver to be calibrated, a feeder for the antenna to be calibrated, and a host unit for a time-frequency transfer receiver to be calibrated. The host unit for a time-frequency transfer receiver to be calibrated outputs a 1PPS second pulse and connects it to a high-precision time interval counter.

[0013] The vector network analyzer is used to measure the group delay (CAB DLY) of the reference antenna feed and the antenna feed under test in the 1.16~1.31GHz and 1.52~1.62GHz frequency bands.

[0014] The high-precision time interval counter collects the deviation data between the output 1PPS of the time-frequency transfer receiver to be calibrated and the standard 1PPS of the time-frequency source allocation module, which is used to calculate the reference delay REF DLY of the time-frequency transfer receiver to be calibrated.

[0015] The entire setup acquires 1-2 days of RINEX observation data under common clock synchronization conditions, and the common clock error is calculated using Dclrinex software. The hardware delay of the time-frequency transfer receiver to be calibrated is solved based on the differential formula.

[0016] The present invention also includes a GNSS time transfer link differential calibration method based on the above-described device, comprising the following steps:

[0017] S1. Sub-item delay pre-calibration: The dual-band group delay (CAB DLY) of the reference antenna feed and the antenna under test feed is measured using a vector network analyzer; 24-hour 1PPS deviation data is continuously collected using a high-precision time interval counter to calculate the reference delay (REF DLY) of the time-frequency transfer receiver to be calibrated; the pre-calibrated hardware delay of the reference time-frequency transfer receiver is retrieved. ;

[0018] S2. Microwave anechoic chamber continuous observation with clock difference: The device is set up with a synchronous time transfer link, a sealed microwave anechoic chamber, a GNSS simulator is loaded with a zero pseudorange error-free simulation scenario, dual receiving antennas with ultra-short baseline synchronously receive simulated GNSS signals, two time-frequency transfer receivers synchronously collect 1~2 days of RINEX observation data, and the experimental environment temperature is stably controlled within ±1℃.

[0019] S3. Common Clock Difference Data Calculation and Hardware Delay Solution: Importing the dual-time transfer link RINEX file and precise ephemeris into Dclrinex software for calculation. Substitute the values ​​into the differential formula to calculate the hardware delay of the time-frequency transfer receiver to be calibrated. Then solve for the total transmission delay of the time transfer link to be calibrated. ;

[0020] The formula for calculating the hardware delay of the time-frequency transfer receiver to be calibrated is as follows:

[0021] (1)

[0022] in and These are the hardware delays of the time-frequency transfer receiver to be calibrated and the reference time-frequency transfer receiver, respectively. The hardware delay of the time-frequency transfer receiver includes the hardware delay of the time-frequency transfer receiver host and the hardware delay of the time-frequency transfer receiver GNSS antenna. The time delay difference between the two clock-time transmission links; and These are the reference delays for the time transfer link to be calibrated and the reference time transfer link, respectively. and These are the antenna feed delays for the time transfer link to be calibrated and the reference time transfer link, respectively.

[0023] Total transmission delay of the time transfer link to be calibrated The calculation formula is as follows:

[0024] (2)

[0025] S4. Layered synthesis standard uncertainty assessment: Calculate the four independent uncertainty components in sequence: differential data solution, antenna feeder delay, reference delay, and reference time-frequency transfer receiver hardware delay. Use the root sum of squares method to synthesize the total standard uncertainty of the calibration result.

[0026] S5. Multi-method cross-validation: The hardware delay of the time-frequency transmission receiver to be calibrated obtained by differential calibration of this method is compared with the measurement results of step-by-step absolute calibration and overall absolute calibration, respectively, to verify whether the three sets of values ​​fall within each other's uncertainty range, and to complete the determination of calibration validity.

[0027] This invention has many beneficial effects.

[0028] 1. Controllable and reproducible experimental environment: The GNSS simulator is used in conjunction with a microwave anechoic chamber to replace the real outdoor satellite signal, actively eliminating interference from the ionosphere, troposphere and multipath effects. There are no time-varying errors in the outdoor environment, and the calibration experiment under the same conditions can be repeated many times, which solves the defect that traditional real signal differential calibration cannot be reproduced and verified.

[0029] 2. High measurement stability: The simulator actively shields various propagation errors, and the microwave anechoic chamber eliminates indoor multipath reflections. The measured standard deviation of the common clock difference observation of Beidou at various frequencies is only 0.14~0.21ns, and the data noise is far lower than that of the outdoor real signal differential calibration.

[0030] 3. Low experimental complexity, suitable for small and medium-sized metrology laboratories: Compared with step-by-step and overall absolute calibration, this invention uses the differential principle to cancel common delay items such as simulator delay and air propagation delay of radio frequency transmission path. It does not require the use of a high-speed oscilloscope to calibrate simulator delay, reduces the need for supporting high-precision test equipment, simplifies the operation process, and supports batch time and frequency transmission receiver hardware delay calibration.

[0031] 4. Excellent calibration accuracy, with uncertainty level comparable to step-by-step absolute calibration: The measured combined uncertainty of BeiDou B1I, B3I, and B1C frequency point calibration is only 0.92~0.94ns, which is comparable to the uncertainty index of step-by-step absolute calibration, meeting the requirements of nanosecond-level time and frequency traceability metrology.

[0032] 5. High reliability of calibration results: Supports cross-validation of differential calibration, step-by-step absolute calibration, and overall absolute calibration data. Multiple verifications avoid the inherent system bias of a single calibration method. It is compatible with the domestic JJF 1403-2013 GNSS time and frequency transfer receiver calibration specification and the BIPM international time and frequency metrology standard.

[0033] 6. High versatility: It is compatible with BeiDou full-frequency point, GPS and other multi-mode GNSS time and frequency transfer receivers, and relies on Dclrinex open source software for calculation, and is compatible with industry-standard RINEX and CGGTTS observation data formats. Attached Figure Description

[0034] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 is a schematic diagram of the overall principle connection of the time transfer link differential calibration device based on the GNSS simulator of the present invention.

[0036] Figure 2 is a schematic diagram of the uncertainty assessment stratification model of the calibration method of the present invention;

[0037] Figure 3 is a schematic diagram of the TLab-TFS-G1 time-frequency transfer receiver used in this embodiment.

[0038] Figure 4 shows the timing variation curve of the common clock difference Δt_CCD at the BeiDou B1I frequency point DC03-TL16 in this embodiment. The horizontal axis is the simplified Julian day, and the vertical axis is the time difference value in ns.

[0039] Figure 5 shows the timing variation curve of the common clock difference Δt_CCD at the BeiDou B3I frequency point DC03-TL16 in this embodiment. The horizontal axis is the simplified Julian day, and the vertical axis is the time difference value in ns.

[0040] Figure 6 shows the timing variation curve of the common clock difference Δt_CCD at the Beidou B1C frequency point DC03-TL16 in this embodiment. The horizontal axis is the simplified Julian day, and the vertical axis is the time difference value in ns. Detailed Implementation

[0041] This embodiment verifies the hardware delay differential calibration of the time-frequency transfer receiver for BeiDou-3 timing frequency points B1I, B3I, and B1C. The entire system hardware selection, experimental environment, step-by-step operation, data calculation, uncertainty synthesis, and multi-method cross-comparison are all fully implemented. All test data and instrument control conditions are original measured data. Technical personnel in the relevant field can completely reproduce the entire calibration process based on the content of this embodiment.

[0042] I. Hardware Device Model and Parameters for Implementation Examples

[0043] Time and frequency source allocation module, with rubidium atomic clock time and frequency allocation unit featuring a constant temperature bath, frequency stability It outputs 8 isolated 10MHz sine reference signals and 10 isolated 1PPS pulse synchronization signals; the rise time of 1PPS is <1 ns, the isolation of each output is >100 dB, and it comes with a matching 50 Ω low-noise synchronization cable of equal length.

[0044] The GNSS signal simulator uses the BeiDou satellite navigation signal simulator and supports BeiDou-2 / 3 signal output; the RF output power is adjustable from -130 dBm to -60 dBm, and the port impedance is 50 Ω; it can independently switch the ionosphere, troposphere, multipath, satellite clock bias, and time-frequency transfer receiver clock bias error model; the carrier elevation supports centimeter-level fine adjustment to achieve pseudorange zeroing; this embodiment enables BeiDou B1I, B3I, and B1C three-channel RF output.

[0045] The microwave anechoic chamber has internal dimensions of 0.8 m × 0.8 m × 1.3 m, with the inner walls fully covered by 200 mm thick polyurethane absorbing cones. It operates in the frequency band of 1.1 GHz to 1.7 GHz, with internal RF reflection attenuation ≥ 60 dB. The metal-sealed door is equipped with conductive foam. Inside, there is one broadband transmitting horn antenna and two GNSS measurement receiving antennas. The distance between the transmitting antenna and the two receiving antennas is 0.6 m, and the distance between the phase centers of the two receiving antennas is 0.25 m, forming an ultra-short baseline configuration.

[0046] The time-frequency transfer receiver includes: a reference time-frequency transfer receiver TLab-TFS-G1 (number TL16) and a time-frequency transfer receiver to be calibrated TLab-TFS-G1 (number DC03); it supports BeiDou three-frequency synchronous tracking, outputs observation files in RINEX3.04 and CGGTTS-V2E formats, and has a 1PPS output jitter of <0.1 ns.

[0047] The RF feeder and adapters include: the reference time transfer link and the time transfer link to be calibrated. The antenna feeder is a 2.5 m coaxial cable with a cable temperature delay coefficient of 0.01 ns / ℃. It comes with two BM / TF and NF / TF adapters. The assembly is uniformly tightened with a torque wrench to avoid deformation introducing time delay errors.

[0048] The vector network analyzer operates in the 1-2 GHz frequency band, is equipped with a dual-port calibration device, and has a group delay measurement resolution of 0.001 ns. The measurement bandwidth is set to 1 kHz, the averaging factor is 16, and each frequency point is sampled 1601 times and the average value is taken.

[0049] The high-precision time interval counter, model SR620, has a single measurement resolution of 0.001 ns, an internal relative error of 0.05 ns, and an inherent resolution error of 0.10 ns; it has a dual-channel input impedance of 50 Ω and supports automatic storage of uninterrupted second pulse difference values.

[0050] The temperature control equipment includes a constant temperature laboratory linkage temperature control system, which controls the ambient temperature at 23±1 ℃ throughout the process, and a temperature and humidity recorder records temperature data every 2 minutes.

[0051] II. Pre-experimental delay pre-calibration procedure

[0052] Step 1: Antenna feed group delay CAB DLY calibration

[0053] The vector network analyzer completed the dual-port full calibration, and then connected the reference antenna feed line and the antenna feed line under test with two experimental adapters in series. The group delay of 1601 uniform frequency points was collected in two frequency bands: 1.16~1.31GHz and 1.52~1.62GHz. The average value of the two frequency bands was taken as the cable delay, and the standard deviation of the measurement was taken as the cable repeatability uncertainty.

[0054] Actual measurement results: The group delay difference between the feeder of the antenna under test and the feeder of the reference antenna in the BeiDou operating frequency band is 14.77 ns, that is... - = 14.77 ns, with a maximum standard deviation of 0.20 ns.

[0055] Step 2: Time-Frequency Transfer Receiver Reference Delay (REF / DLY) Calibration

[0056] The time-frequency distribution module outputs a standard 1PPS, which is connected in two ways to the 1PPS input ports of two time-frequency transfer receivers and the A ports of two SR620 counters. The 1PPS outputs of the two time-frequency transfer receivers are connected to the B ports of the two SR620 counters respectively. The counters are set to store data at a 1-second interval, continuously collecting 24-hour pulse difference data. After removing unstable data from the first hour of power-on, the average value is taken, and the standard deviation of the data is used as the measurement repeatability uncertainty.

[0057] Actual test results: = 71.76 ns, measurement standard deviation 0.427 ns; =71.24 ns, measurement standard deviation 0.419 ns.

[0058] Step 3: Retrieve the reference time transfer link and the known time-frequency transfer receiver hardware delay.

[0059] The reference time-frequency transfer receiver TL16 was pre-calibrated through overall absolute calibration. The hardware delay values ​​of the time-frequency transfer receiver for each BeiDou frequency point are: 24.20 ns for B1I frequency point, 27.42 ns for B3I frequency point, and 31.37 ns for B1C frequency point. The combined standard uncertainty corresponding to this calibration result was retrieved synchronously.

[0060] III. Continuous Observation of Simulated Signal Common Clock Difference in Microwave Dark Box

[0061] Complete all RF and synchronization cable wiring checks, close the microwave anechoic chamber door, and shut down the laboratory wireless transmitting equipment; create a static carrier scene in the GNSS simulator, fine-tune the elevation to zero the satellite-carrier pseudorange, disable all error models one by one, and uniformly adjust the output power of the BeiDou three-frequency signals to -110 dBm; synchronously set the RINEX sampling interval to 30 s for both time-frequency transfer receivers and synchronously start observation data storage; the entire system runs continuously for 2 days, corresponding to the Simplified Julian Day (MJD) interval 60041.0~60043.0; after the observation is completed, export the complete RINEX observation file and the corresponding BeiDou precise ephemeris.

[0062] IV. Common Clock Error Calculation and Hardware Delay Calculation

[0063] Import the RINEX file and precise ephemeris into the BIPM open-source Dclrinex software to calculate the Δt_CCD for each frequency point: B1I mean 22.07 ns, standard deviation 0.21 ns; B3I mean 16.10 ns, standard deviation 0.16 ns; B1C mean 17.11 ns, standard deviation 0.14 ns.

[0064] Substitute each known parameter into equation (1) to calculate. :

[0065] B1I frequency: 32.02 ns; B3I frequency: 29.27 ns; B1C frequency: 34.23 ns.

[0066] Will Substituting CAB DLY and REF DLY into equation (2) yields the total transmission delay of the complete time transfer link to be calibrated. This serves as the final calibration output.

[0067] V. Assessment of Stratified Uncertainty (Taking B1I Frequency as an Example)

[0068] The uncertainty of this invention is divided into four uncorrelated components, and the total standard uncertainty is synthesized using the root sum of squares method.

[0069] 1. Uncertainty in solving difference data Including CCD observation repeatability of 0.21 ns, reference time-frequency transfer receiver positioning error of 0.05 ns, and the positioning error of the time-frequency transfer receiver to be calibrated of 0.05 ns, the calculation yields... =0.2216ns;

[0070] 2. Antenna feeder delay calibration uncertainty The calculation includes cable repeatability (0.20 ns), VNA dual-port calibration (0.20 ns), temperature fluctuation (0.12 ns), two adapters combined (0.20 ns), and cable deformation (0.15 ns). =0.3961 ns;

[0071] 3. Reference delay calibration uncertainty Including measurement repeatability of 0.427 ns, counter resolution of 0.10 ns, and counter relative error of 0.05 ns, the calculated values ​​are... =0.4414 ns;

[0072] 4. Uncertainty of hardware delay calibration of reference time-frequency transfer receiver The combined uncertainty of the hardware delay precalibration of the reference time-frequency transfer receiver is 0.69 ns.

[0073] Total combined standard uncertainty The same calculation logic was used for the B3I and B1C frequencies, with combined uncertainties of 0.92 ns and 0.93 ns, respectively.

[0074] VI. Cross-validation using multiple methods

[0075] Simultaneously employing the differential calibration, step-by-step absolute calibration, and overall absolute calibration of this invention, comparative tests were conducted on the hardware delay of the same time-frequency transfer receiver under test. The three sets of calibration results and uncertainties are shown in the table below:

[0076] B1I 32.02 0.94 33.51 0.95 33.26 0.69 B3I 29.27 0.92 30.70 0.92 30.52 0.69 B1C 34.23 0.93 35.17 0.93 34.58 0.69

[0077] The criterion for mutual agreement is that the difference between the calibration values ​​from the two methods is less than the sum of their uncertainties; in this case, the results are considered to be in agreement. The measured values ​​from all three schemes fall within each other's uncertainty coverage range, proving that the calibration results of this invention have no significant systematic bias and that the measurement results are reliable.

[0078] This embodiment takes the BeiDou three-frequency time-frequency transfer receiver as an example. The technical solution of this invention can be extended to the calibration of other GNSS time transfer links such as GPS and GLONASS. It only requires enabling the corresponding constellation signal in the GNSS simulator and adjusting the test frequency band of the vector network analyzer. The differential calculation model and uncertainty assessment process do not need to be modified, and it has wide applicability.

[0079] It should be noted that the above embodiments of the present invention are merely illustrative examples and are not intended to limit the scope of protection of the present invention. Any modifications, substitutions or improvements made by those skilled in the art to the technical solutions of the present invention without departing from the core ideas of the present invention should fall within the scope of protection of the present invention.

Claims

1. A time-transfer link differential calibration device based on GNSS analog signals, characterized in that: It includes a time and frequency source allocation module, a GNSS signal simulator, a microwave anechoic chamber, a reference time transfer link, a time transfer link to be calibrated, a high-precision time interval counter, and a vector network analyzer; The reference time transfer link includes a reference time and frequency transfer receiver GNSS antenna, a reference antenna feeder, and a reference time and frequency transfer receiver main unit, which stores the known hardware delay of the reference time and frequency transfer receiver. , The reference time-frequency transfer receiver is a time-frequency transfer receiver that has been pre-calibrated to absolute accuracy; The time transfer link to be calibrated includes a GNSS antenna for the time-frequency transfer receiver to be calibrated, a feed line for the antenna to be calibrated, and a host unit for the time-frequency transfer receiver to be calibrated. The host unit outputs a 1PPS pulse per second, which is then connected to a high-precision time interval counter. The time-frequency transfer receiver to be calibrated is the time-frequency transfer receiver under test. The time-frequency source allocation module outputs a unified 10MHz frequency signal and a 1PPS second pulse signal, which are respectively connected to the GNSS signal simulator, the reference time-frequency transfer receiver, the time-frequency transfer receiver to be calibrated, and the high-precision time interval counter. The GNSS signal simulator is configured with a static carrier scene, which can turn off all error models such as ionospheric delay, tropospheric delay, multipath effect, satellite clock error, and time-frequency transfer receiver clock error, and adjust the pseudorange between the satellite and the carrier to zero. The simulator's radio frequency output is connected to the transmitting horn antenna inside the microwave anechoic chamber. The microwave anechoic chamber is equipped with a transmitting horn antenna, a reference time-frequency transfer receiver GNSS antenna and a time-frequency transfer receiver GNSS antenna to be calibrated, arranged with an ultra-short baseline. The reference time-frequency transfer receiver GNSS antenna and the time-frequency transfer receiver GNSS antenna to be calibrated are respectively connected to the radio frequency input ports of the reference time transfer link and the time transfer link to be calibrated through independent antenna feed lines. The microwave anechoic chamber shields external electromagnetic interference and eliminates radio frequency multipath reflections in a sealed state. The vector network analyzer is used to measure the group delay (CAB DLY) of the reference antenna feed and the antenna feed under test in the 1.16~1.31GHz and 1.52~1.62GHz frequency bands. The high-precision time interval counter collects the deviation data between the output 1PPS of the time-frequency transfer receiver to be calibrated and the standard 1PPS of the time-frequency source allocation module, which is used to calculate the reference delay REF DLY of the time-frequency transfer receiver to be calibrated. The entire device acquires RINEX observation data for 1-2 consecutive days under common clock synchronization conditions. The common clock difference between the reference time transfer link and the time transfer link to be calibrated is calculated using Dclrinex software. The hardware delay of the time-frequency transfer receiver to be calibrated is solved based on the differential formula.

2. The time-transfer link differential calibration device based on GNSS analog signals according to claim 1, characterized in that, The time-frequency transfer receiver is a multi-mode GNSS time-frequency transfer receiver that supports BeiDou-2 / 3 and GPS signal tracking, and can output RINEX and CGGTTS standard observation files.

3. The time-transfer link differential calibration device based on GNSS analog signals according to claim 1, characterized in that, The high-precision time interval counter is model SR620, with an inherent resolution error of 0.10 ns and an internal relative error of 0.05 ns.

4. The time-transfer link differential calibration device based on GNSS analog signals according to claim 1, characterized in that, The two receiving antennas inside the microwave anechoic chamber are spaced ≤0.5 m apart, forming an ultra-short baseline configuration.

5. A GNSS time transfer link differential calibration method based on the device described in any one of claims 1 to 4, characterized in that, Includes the following steps: S1. Pre-calibration of component delays: The dual-band group delay (CAB DLY) of the reference antenna feed and the antenna feed to be calibrated is measured using a vector network analyzer; 24-hour 1PPS deviation data is continuously collected using a high-precision time interval counter to calculate the reference delay (REF DLY) of the time-frequency transfer receiver to be calibrated; the pre-calibrated hardware delay of the reference time-frequency transfer receiver is retrieved. ; S2. Microwave anechoic chamber continuous observation with clock difference: A sealed microwave anechoic chamber is used. A GNSS simulator is loaded with a zero pseudorange error-free simulation scenario. Dual receiving antennas synchronously receive simulated GNSS signals. Two time-frequency transfer receivers synchronously and continuously collect 1~2 days of RINEX observation data. The experimental environment temperature is stably controlled at ±1 ℃. S3. Common Clock Error Data Calculation and Hardware Delay Solution: Import the dual-time transfer link RINEX file and precise ephemeris into Dclrinex software to calculate the common clock error. Substitute into the following formula to calculate the hardware delay of the time-frequency transfer receiver to be calibrated. : in and These are the hardware delays of the time-frequency transfer receiver to be calibrated and the reference time-frequency transfer receiver, respectively, including the hardware delay of the time-frequency transfer receiver host and the hardware delay of the GNSS antenna of the time-frequency transfer receiver to be calibrated. The time delay difference between the two clock-time transmission links; and These are the reference delays for the time transfer link to be calibrated and the reference time transfer link, respectively. and These are the antenna feed delays for the time transfer link to be calibrated and the reference time transfer link, respectively. Based on Solve for the total transmission delay of the time transfer link to be calibrated. ; The hardware delay of the time-frequency transfer receiver to be calibrated includes the hardware delay of the host of the time-frequency transfer receiver to be calibrated and the hardware delay of the GNSS antenna of the time-frequency transfer receiver to be calibrated. S4. Layered Combined Standard Uncertainty Assessment: Calculate the four types of uncertainty components: differential data solution, antenna feeder delay, reference delay, and reference time-frequency transfer receiver hardware delay, and combine the root of square to obtain the total standard uncertainty U. S5. Multi-method comparison and verification: The time-frequency signal to be calibrated obtained by this method is transferred to the receiver hardware delay. The calibration validity is determined by comparing the results with those of stepwise absolute calibration and overall absolute calibration.

6. The GNSS time transfer link differential calibration method according to claim 5, characterized in that, The antenna feeder delay uncertainty in step S4 includes cable measurement repeatability, VNA dual-port calibration 0.20 ns, ambient temperature fluctuation 0.12 ns, single adapter 0.1 ns, and cable deformation 0.15 ns.

7. The GNSS time transfer link differential calibration method according to claim 5, characterized in that, The reference delay uncertainty in step S4 includes the reference delay measurement repeatability, the SR620 resolution error introducing 0.10 ns, and the SR620 relative error introducing 0.05 ns.

8. The GNSS time transfer link differential calibration method according to claim 5, characterized in that, The uncertainty items in the differential data solution in step S4 include the repeatability of CCD observation data, the positioning error of the reference time-frequency transfer receiver (0.05 ns), and the positioning error of the time-frequency transfer receiver to be calibrated (0.05 ns).

9. The GNSS time transfer link differential calibration method according to claim 5, characterized in that, In step S2, the sampling interval of the RINEX file in the time-frequency transfer receiver is set to 30 s.

Images