RU2769113C1 - Method of correcting sampling error of ranging code - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: invention relates to radio navigation, namely to generation, reception and processing of broadband radio navigation signals, and can be used to generate a ranging code in hardware and software simulators and signal receivers of radio navigation systems of satellite and ground radio navigation, including precision equipment of metrological class, receivers located on spacecraft, time synchronization equipment according to signals of global navigation satellite systems. Distance measuring code sampling error correction method consists in calculating, for each transition between PRS symbols, the sum of the value of the deviation of the position of the current transition between the symbols of the discretized PRS from the required position and the sum of deviations calculated for previous transitions, and shifting the current transition between the PRS symbols in time by one sampling cycle towards the advance. Shift of the current transition between the PRS symbols is performed if the obtained total value of the deviation value is positive. Calculation of subsequent sums of deviations of positions of transitions between symbols of PRS is performed taking into account performed correction of position of current transition between symbols of PRS.

EFFECT: reduced sampling error of the rangefinder code due to reduced change in the average value of deviations of transitions between symbols of the pseudorandom sequence (PRS), calculated on a time interval equal to the PRS period or different from the PRS period when the delay value is changed.

1 cl, 10 dwg

Description

The invention relates to the field of radio navigation, namely to the formation, reception and processing of broadband radio navigation signals and can be used to generate a ranging code in hardware and software simulators and signal receivers of radio navigation systems of satellite and terrestrial radio navigation, including precision equipment of the metrological class, receivers located on spacecraft (SC), time synchronization equipment based on the signals of global navigation satellite systems (GNSS).

In modern radio navigation systems, the position of the consumer (user) of the navigation system is determined on the basis of the distances between the consumer and the reference stations (spacecraft) of the system emitting navigation signals. Distances are determined by measuring the propagation time of navigation signals from reference stations to the consumer's receiver. As navigation signals, as a rule, broadband signals are used, modulated by ranging codes - pseudo-random sequences (PRS) of zeros and ones.

To measure the propagation time (delay) of the signal, the receiver generates a local copy of the ranging code, the delay of which is continuously adjusted to the delay of the input signal using a tracking system. The output signal of the tracking system discriminator is a signal proportional to the difference in the output values of two correlators, the reference ranging codes of which are shifted in time relative to the current delay estimate in the direction of advance and delay by an amount usually not exceeding half the duration of the PRS symbol. Essentially, the discriminator determines the time offset between the PRS symbol transitions of the input signal and the PRS symbol transitions generated at the receiver. Thus, information about the delay of the input signal is extracted from the time positions of the transitions between PRS symbols. The control of the delay of the generated ranging code is carried out from the output of the loop filter of the tracking system by changing the frequency of the PRS symbols, or by directly setting the required delay in the controlled ranging code generator.

Obviously, the accuracy of measuring the delay of the received signal directly affects the accuracy of determining the position of the consumer. However, in the digital implementation of algorithms for processing received navigation signals, which has become widespread due to a number of reasons, the finite sampling rate leads to a discretization error in the ranging code, and, as a result, to an error in measuring the propagation time of the navigation signal.

When discretizing the ranging code, the positions of the transitions between the symbols of the PRS are shifted and, as a result, the zero shift of the discrimination characteristic occurs. In the general case, the offset is random, uniformly distributed in the range from zero to the duration of the sampling period, depends on the specific value of the delay of the PRS, the PRS itself, and the sampling frequency.

The occurrence of offset transitions between PR symbols during sampling is illustrated in FIG. 1. As an example, the timing diagrams of a part of the PSP are given, four symbols long, with two transitions. Used in FIG. 1 designations: d-value of the PSP symbol, T-duration of the PSP symbol, T _d - sampling period. The dots (circles) denote samples of the discrete PRS. The solid line corresponds to the formed discrete memory bandwidth. Dashed vertical lines mark the positions of the boundaries of the symbols of the ideal (not sampled) PRS. e is the value of the offset of the transition between the symbols of the sampled PR with respect to the ideal PR. Timing diagrams are given for zero delay (upper diagram) and delay equal to τ (lower diagram). From FIG. 1 it can be seen that the average value of the deviations of the transitions between the PR symbols depends on the delay of the PR. As the duration of the PRS (the number of transitions between the symbols of the PRS) increases, the transition offsets are averaged, and the resulting error in determining the delay decreases. However, in a number of tasks, the final value of the error remains unacceptable.

When measuring a varying delay (and the presence of a Doppler shift in the frequency of the input navigation signal), the error is also averaged. However, for small and zero DFS values, the effect of averaging is small or absent.

The need for high-precision formation of the ranging code and measurement of its delay arises in: measuring stations of the ground segment of the GNSS; equipment of consumers of navigation systems located on spacecraft; metrological class devices (simulators and receivers of navigation signals used for calibration and verification of equipment); time synchronization equipment based on GNSS signals; problems of measuring the distance between objects of aviation and space technology.

Reducing the sampling error of the ranging code by increasing the sampling frequency is limited primarily by such factors as: the complexity of implementing high-frequency signal processing (computing resources), power consumption, heat dissipation, and weight and size characteristics.

It should be noted that in the case of a multiplicity of the sampling frequency of the repetition rate of the PRS symbols, the sampling error significantly increases (article by Dennis M. Akos, Marco Pini, Effect of Sampling Frequency on GNSS Receiver Performance, NAVIGATION: Journal of The Institute of Navigation Vol.m53, No. 2, Summer 2006).

There is a known method for reducing the sampling error of the ranging code, which consists in calculating the zero offset of the discrimination characteristic based on the analysis of the positions of the ranging code samples and the subsequent correction of pseudorange measurements by the value of the calculated offset (article by Xiaojun Jin, Ning Zhang, Kan Yang et al., PN Ranging Based on Noncommensurate Sampling : Zero-Bias Mitigation Methods, IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS VOL. 53, NO. 2 APRIL 2017). Another method proposed in the article to reduce the sampling error of the ranging code is the choice of special values of the sampling frequency, at which the error is compensated when subtracting values at the outputs of correlators with lagging and leading reference signals. The disadvantage of the error correction method proposed in the article, based on the calculation of the zero offset of the discrimination characteristic, is a large amount of calculations. The choice of a special value for the sampling frequency is difficult due to the practical aspects of the implementation of the equipment (synthesizer tuning, frequency plan), it is impossible in the widespread practice of sampling signals from several navigation systems (GLONASS, GPS, Galileo) with a single analog-to-digital converter (ADC). In addition, as the authors of the article note, in order to apply the method they propose in the case of a small time mismatch between the leading and lagging channels of the discriminator (a simple and widely used technique for combating multipath signal propagation), a high sampling rate is required. Both methods proposed in the article are not applicable for implementation in navigation signal simulators.

Another way to correct the sampling error of the ranging code is to pre-calculate the sampling error for all possible delay values of the ranging code in the range from zero to the duration of the sampling cycle (patent EP 3 483 632 A1, WO 2019/091769 A1, article J. - M. Sleewaegen, W De Wilde, Zero-Doppler Pseudorange Biases, 2018, Septentrio, Belgium). The calculated error value is filtered with a low-pass filter having the same characteristics as the main loop filter of the delay tracking system. The filtered error value is used to correct the pseudorange measurements. The disadvantage of this method is the need to implement an additional filter. In addition, the method is not applicable for implementation in simulators of navigation signals.

There is also a known method for reducing the sampling error of the rangefinder code, which consists in increasing the sampling rate of the input signal using interpolation (article by P.V. Sharshavin, A.S. Kondratiev, A.V. Grebennikov, The use of digital registration to improve the accuracy characteristics of pseudorange measurement by signals satellite radio navigation systems GLONASS/GPS, 2012 Bulletin of the Siberian State Aerospace University named after Academician M. F. Reshetnev, article by P. V. Sharshavin, A. V. Grebennikov, Study of the method of post-processing of satellite radio navigation systems signals with upsampling, 2017, NPP LLC "Autonomous Aerospace Systems - GeoService", Institute of Engineering Physics and Radioelectronics, Siberian Federal University, Krasnoyarsk, https://uavsiberia.com/en/news/issledovanie-metoda-postobrabotki-signalov-sputnikovykh-radionavigatsionnykh-sistem -s-povysheniem-chastoty-diskretizatsii/). Interpolation is performed using a sample rate expander and an interpolator filter. The formation of the reference ranging code in the receiver and the calculation of the correlation values is carried out at a high sampling rate. The disadvantage of this method, as noted by the authors of the article, is the need for high computational costs. This method can be used in the mode of post-processing of records of navigation signals in a software receiver, however, its application in real-time equipment is difficult. The proposed method also cannot be used in navigation signal simulators.

There is also known a method for correcting the sampling error of the ranging code, which consists in limiting the spectrum of the ranging code signal (article by P.V. Sharshavin, A.S. Kondratiev, A.V. Grebennikov, Reducing the sampling error of the pseudo-random sequence of the ranging code by limiting the spectrum, Communication systems and radio navigation: collection of abstracts / scientific editor V.F. Shabanov, responsible for the issue A.Yu. Strokova. - Krasnoyarsk: JSC "NPP "Radiosvyaz", 2015. -355 p.). According to the description given in the article, spectrum limitation is implemented by adding an additional block to the PRS formation circuit with a phase accumulator, which detects the moments of the PRS transition and reproduces the transient process at these moments. Reducing the sampling error of the ranging code is achieved due to the fact that when reproducing the transient process, the state of the accumulator of the phase of the PRS symbol frequency generator at the moments of the transition between the PSS symbols, which carries information about the true position of the transition, is taken into account. Reproduction of transient processes in a ranging code signal having a limited spectrum is possible under the condition of multilevel quantization of this signal. The disadvantage of this method is the complication of implementation - an increase in the number of levels of quantization of the ranging code signal compared to the traditionally used binary quantization of this signal. An increase in the number of quantization levels entails a complication of the correlator (there is a need for storage devices, a multi-bit multiplier).

Closest to the claimed invention is a method for correcting the ranging code sampling error, which consists in shifting the temporal positions of transitions between PSS symbols (article by Vinh T. Tran, Nagaraj C. Shivaramaiah, Thuan D. Nguyen, Eamonn P. Glennon, Andrew G. Dempster, GNSS receiver implementations to mitigate the effects of commensurate sampling frequencies on DLL code tracking, GPS Solutions, 2018) by a random variable, taken as a prototype. In accordance with the prototype, in order to reduce the sampling error of the ranging code, the positions of the transitions between the symbols of the SRP are shifted by a small random value within the duration of the sampling cycle. At the same time, the article proposes to use a random value, or a value that varies according to a harmonic or linear law, as the displacement value. In the case of a multiplicity of the sampling rate and the repetition rate of the PRS symbols, the method described in the article allows to reduce the discretization error of the ranging code. However, with non-multiple sampling rate and PRS symbol frequency, since the value of the main phase accumulator at each change of PRS symbols is random, adding an additional value, random, or changing according to a harmonic or linear law, does not eliminate the error. The disadvantage of the prototype is a significantly large amount of error remaining after correction.

The purpose of this invention is to reduce the discretization error of the ranging code, to reduce the error in reproducing the delay of the ranging code in radio navigation signals generated by simulators, and to reduce the delay measurement error by radio navigation signal receivers.

This goal is achieved by changing the temporal positions of the transitions between the PRS symbols during the formation of the ranging code, namely, for each transition between the PSP symbols, the sum of the deviation of the position of the current transition between the symbols of the sampled PRP from the true position and the sum of the deviations calculated for previous transitions are calculated, and, if the obtained total value of the deviation value is positive, then the current transition between the PSP symbols is shifted in time by one sampling cycle in the direction of advance, while the calculation of the subsequent sums of deviations of the positions of the transitions between the PSP symbols is performed taking into account the correction of the position of the current transition between PSP symbols.

The technical result of performing the above actions is to reduce the discretization error of the ranging code due to a decrease in the change in the average value of deviations of transitions between PSP symbols, calculated on a time interval equal to the PSP period, or different from the PSP period when the delay value changes. Timing diagrams illustrating the transition shift between PR symbols are shown in FIG. 2. Bold dotted lines show the position of the transition before correction.

The proposed method for reducing the sampling error of the ranging code and the method used in the prototype involve performing actions directly with the transitions of the ranging code, in contrast to the analogues given in the description of the invention, in which the delay measurement error is corrected. Due to this, the methods proposed and used in the prototype can be used to reduce the sampling error of the ranging code generators of both receivers and simulators of radio navigation signals.

The difference between the claimed invention and the prototype is that the displacement of the positions of transitions between PSP symbols is performed not randomly, but taking into account the deviation of the current and previous PSP symbols from the true value.

In FIG. 1 shows the timing diagrams explaining the shift of transitions between PSS symbols (the occurrence of a discretization error of the ranging code).

In FIG. 2 shows the timing diagrams after performing the correction of the ranging code sampling error.

In FIG. Figure 3 shows an example of a block diagram of a PRS generator that implements the proposed method for correcting the sampling error of a ranging code.

In FIG. Figure 4 shows a block diagram of the algorithm for the software implementation of the proposed method for correcting the discretization error of the ranging code.

In FIG. Figure 5 shows graphs of cross-correlation functions without correction of the discretization error of the ranging code, obtained by modeling.

In FIG. 6 shows graphs of cross-correlation functions using the proposed method for correcting the discretization error of the ranging code obtained by modeling.

In FIG. Figure 7 shows the graphs of the ranging code delay measurement error versus the input signal delay value without correcting the ranging code sampling error, obtained by simulation.

In FIG. Figure 8 shows the graphs of the ranging code delay measurement error versus the value of the input signal delay with correction of the ranging code sampling error, obtained by simulation.

In FIG. Figure 9 shows a graph of the ranging code delay measurement error without correcting the experimentally obtained ranging code sampling error.

In FIG. Figure 10 shows a graph of the rangefinder delay measurement error with the rangefinder code sampling error correction obtained experimentally.

An example of a block diagram of a device that implements the proposed method for correcting the ranging code sampling error is shown in Fig. 3.

The formation of the clock frequency of the PRP symbols (f _T ) is performed using the accumulator of phase 1, containing the adder 2 and the register of the accumulator of phase 3. N is the capacity of the phase accumulator. The required symbol rate is set using the frequency code (K _h ) supplied to the second input of the adder 2. Unit 4 detects transitions of the most significant bit of the phase accumulator from the logic level "1" to the logic level "0". MSB - most significant bit (most significant bit). According to the signal of the clock frequency of the symbols of the SRP (from the output of block 4), the shift register with linear feedback (LFSR, block 5) is shifted or the transition to the next memory cell in the case of tabular formation of the SRP. In a traditional ranging code generator circuit, the output of block 5 is the output of the circuit.

To implement the proposed method for reducing the ranging code sampling error, the circuit (Fig. 3) is supplemented with block 6, which calculates the sum (block 12) of the deviation of the position of the current transition between the symbols of the sampled PRS from the true position and the sum of the deviations calculated for previous transitions. To calculate the amount of deviation on the current transition between PRS symbols from the true value, the value of the PRS clock phase accumulator is used. The value of the sum of the deviations of the transitions (register 13) is updated if (the condition is implemented using a signal at the clock enable input - CE) the next symbol d _k differs from the previous d _k-1 (block 9). Register 7 serves to delay the generated SRP symbols by one sampling cycle. If the value of the sum of deviations of the positions of transitions between PSP symbols from the true value is positive (block 14), then an undelayed PSP symbol is output using the selection element (block 8) (the position of the transition between PSP symbols is corrected). If the error value accumulated in the error accumulator (block 9) is negative, the position of the transition between the PSP symbols is not corrected and the PSP symbol delayed by one sampling cycle is output. With each change of the SRP symbol, the value of the error accumulator 13 is updated by the amount of deviation of the current transition from the required value, taking into account whether the correction is performed or not (block 10). The register (block 11) is designed to take into account the delay that occurs in the LFSR.

An example of a software implementation of the method for correcting the sampling error of the ranging code is an algorithm, the block diagram of which is shown in Fig.4.

The following notation is used in the block diagram: e is the sum of the deviation values of transitions between PSP symbols; d _k - current PSP symbol; d _k-1 - previous PSP symbol; acc - the state of the accumulator of the phase of the clock frequency generator of the PSP at the current time; increment - increment of the accumulator of the phase of the PSP clock generator (frequency code); forbid - flag for prohibiting the formation of a clock frequency pulse (performs a function similar to blocks 5, 10 in figure 3 in hardware implementation); MSB is the most significant bit of the phase accumulator.

The feasibility of the invention is confirmed by the results of software simulation and experiment using a debug board containing a Xilinx Zynq-7000 system-on-a-chip, reference frequency synthesizers, a mixer, a digital-to-analog converter, and a high-speed digital oscilloscope.

The simulation was carried out for the GLONASS-M signal, PSP of standard accuracy. The input signal was formed at a sampling rate of 10 GHz. Next, the input signal was filtered with a bandpass filter, with characteristics similar to those used in radio navigation receivers, and decimated to reduce the sampling rate to 100 MHz. The signal delay measurement error was estimated from the zero shift of the discrimination characteristic. The simulation was carried out for two variants of the formation of the SRP: without using the method of correcting the discretization error of the ranging code and using the proposed method. In FIG. Figure 5 shows the graphs of the cross-correlation functions of the input and reference signals obtained during the simulation, calculated using two correlators, the reference ranging codes of which are detuned by an amount equal to several sampling cycles. In FIG. Figure 6 shows the graphs of the VKF obtained using the method for correcting the discretization error of the ranging code. Curve A _L (Fig. 5, 6) corresponds to the correlator with an advanced reference ranging code, curve A _R corresponds to the correlator with a lagging reference ranging code. The horizontal axis shows the time difference between the delay of the input signal and the delay of the reference signal, normalized to the duration of the sampling period T _d . As can be seen from the graphs, the application of the method of reducing the discretization error of the ranging code leads to a decrease in the error of the obtained cross-correlation functions. The dependence of delay measurement errors on the value of the delay of the input signal is shown in Fig.7 (without correction of the sampling error of the ranging code), Fig.8 (corrected by the proposed method).

To confirm the possibility of achieving a technical result (reducing the ranging code sampling error), an experiment was conducted to generate a real GLONASS navigation signal of standard accuracy in the L1 frequency range using the proposed method for correcting the ranging code sampling error.

The signal was formed using a radio navigation signal simulator implemented on a debug board containing a Zynq-7000 system-on-a-chip, reference frequency synthesizers, a mixer, and a digital-to-analog converter.

The digital part of the simulator was implemented in a user-programmable gate array of the Zynq-7000 chip. One of the navigation signal generation channels contained a PRS clock frequency generator, made in accordance with FIG. 3, i.e. using the proposed method for correcting the sampling error of the ranging code, the other channel of the simulator contained the traditional version of the generator. The processor part of the Zynq-7000 was used to control the digital part of the simulator, including the calculation of frequency codes, the initial values of the phase accumulators of the reference oscillation generators, and the PRS clock frequency based on the specified delay of the generated signal.

The formation of the navigation signal was performed at a sampling frequency of 190 MHz, then, using a mixer, the signal was transferred to a frequency of 1602 MHz. The rate of change of the navigation signal delay set in the simulator was set to 0.1 mm/s. Such a low rate of change of the delay of the navigation signal made it possible to obtain the values of errors in the measurement of the delay for all possible positions of the transitions between the PRS symbols. At the same time, it took 4 hours and 23 minutes to change the delay by an amount equal to the duration of the sampling cycle. During one session of the experiment, the total delay change was about three sampling cycles.

Estimation of the delay of the generated navigation signal was carried out by software analysis of the signal records obtained using a high-speed digital oscilloscope operating at a frequency of 5 GHz. To determine the delay measurement error, the differences in code (measured by the delay of the ranging code) and phase (measured by the phase of the carrier wave) pseudoranges were used. In FIG. 9 shows a graph of the error in measuring the code pseudorange without correcting the sampling error of the ranging code. Fig. 10 - code pseudo-range measurement error when using the proposed correction method in the simulator. The results of the experiment indicate the effectiveness of the proposed method for correcting the discretization error of the ranging code.

Claims (1)

A method for correcting the sampling error of a ranging code, which consists in shifting the temporal positions of transitions between symbols of a pseudo-random sequence (PRS), characterized in that for each transition between PRS symbols, the sum of the deviation of the position of the current transition between the symbols of the sampled PRS from the required position and the sum of the deviations calculated for previous transitions, and if the resulting total value of the deviation value is positive, then the current transition between the PSP symbols is shifted in time by one sampling cycle towards the lead, while the calculation of the subsequent sums of deviations of the positions of the transitions between the PSP symbols is performed taking into account the performed correcting the position of the current transition between PSP symbols.

Images