Timing synchronization system and method based on chirp signal

Abstract

The invention discloses a chirp signal-based timing synchronization system and method, belongs to the field of wireless communication, and can be used for large-scale timing drift compensation in a low-noise scene. The system comprises a timing error detection module, a numerical control oscillator module and an interpolation filter module which are cascaded, the timing error detection module is used for removing chirp single-tone spectrum peak positioning and timing errors, and the numerical control oscillator module is used for mapping the timing errors to stepping factors, updating phase register values and determining integer and decimal interpolation sampling points; the interpolation filter module resamples the chirp symbols by using interpolation sampling points to realize timing synchronization; the method comprises the steps of firstly initializing parameters, then calculating a timing error value of an upper chirp leading signal through a timing error detection module, then calculating a resampling parameter through a numerical control oscillator, and finally obtaining a timing synchronization result after the upper chirp signal is resampled through an interpolation filter module. According to the invention, the timing synchronization anti-noise performance and the estimation precision of the timing error value are improved.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a timing synchronization system and method based on chirp signals, which can be used for large-scale timing drift compensation in low-noise scenarios. Background Art

[0002] Timing synchronization is a key step to ensure accurate signal demodulation in digital communication systems. It aligns the received wireless signal with the local sampling rate so that the sampling time is accurately aligned with the symbol boundary, thereby avoiding inter-symbol interference (ISI) and improving system performance. The timing synchronization process is to estimate the timing deviation between the received signal and the local synchronization sequence and adjust the sampling time of the signal so that the sampling point of the received signal is aligned with the local sampling rate, thereby reducing sampling errors and inter-symbol interference. For a complete timing synchronization system, it mainly includes the estimation of timing error parameters and the feedback correction process of the error. In recent years, the research on timing synchronization algorithms has mainly revolved around the Gardner algorithm, which calculates the timing error of the received signal itself, maps the estimated timing error to the resampling parameter, and adjusts the local sampling clock through the calculated resampling parameter, ultimately achieving accurate timing synchronization.

[0003] For example, the patent document with publication number CN113162713A and titled "Variable Symbol Rate Timing Recovery Method and System Based on Gardner Algorithm" discloses a timing synchronization system based on the Gardner algorithm, which includes a signal receiver module, a cubic interpolator module, a Gardner timing error detection module, a loop filter module, and an interpolation estimation module. The invention interpolates the baseband signal using the sampling decision conditions generated by the interpolation estimator; the timing error is obtained using the Gardner algorithm module under the triggering of the peak moment and transition value moment generated by the interpolation estimator; the timing error is loop-filtered, and the sampling decision conditions and the peak moment and transition value moment generated by the interpolation estimator are updated using the interpolation estimator. However, the invention uses the Gardner algorithm in the timing error detection step. Since the algorithm has significant noise interference, it is difficult for the algorithm to effectively distinguish between signals and noise, resulting in an increase in timing error, especially in a low signal-to-noise ratio environment. Reduced accuracy. Summary of the invention

[0004] In order to overcome the shortcomings of the above-mentioned prior art, the purpose of the present invention is to provide a timing synchronization system and method based on a chirp signal, by using a local de-chirp signal to solve the timing error parameters of the received chirp symbol, calculating the resampling parameters of the received signal through the timing error parameters, and using an interpolation filter to resample the received signal, thereby solving the problem of low synchronization accuracy in a low signal-to-noise ratio environment in the existing timing synchronization technology.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is:

[0006] A timing synchronization system based on chirp signals, comprising a cascaded timing error detection module, a digitally controlled oscillator module and an interpolation filter module;

[0007] The timing error detection module comprises a cascaded de-chirp signal peak position operation submodule and a timing error parameter calculation submodule;

[0008] The de-chirp signal peak position calculation submodule comprises a cascaded de-chirp unit, an FFT unit, a peak position detection unit, and a peak position correction unit, wherein the de-chirp unit conjugates and multiplies each up-chirp symbol with a local de-chirp symbol to obtain de-chirp symbols; the FFT unit performs spectrum calculation on the de-chirp symbols to obtain spectrum data of each de-chirp symbol; the peak position detection unit detects the maximum value position of the spectrum data modulus of each de-chirp symbol, and simultaneously extracts the spectrum data on the left and right sides including the maximum value; the peak position correction unit corrects the position of the maximum value of the frequency domain data of each de-chirp symbol through the maximum value spectrum data and the spectrum data on the left and right sides to obtain the spectrum data peak correction position;

[0009] The timing error parameter calculation submodule includes a cascaded least square fitting unit and a timing error parameter mapping unit, wherein the least square fitting unit performs a line segment fitting on the chirp symbol sequence number and the spectrum data peak correction position relationship curve to obtain a timing error value of a chirp symbol; the timing error parameter mapping unit calculates the timing error value within the scale of a single complex envelope signal through the timing error value of a chirp symbol.

[0010] The FFT unit includes a plurality of cascaded butterfly operation structures, each butterfly operation structure includes parallel arrangement A butterfly operation structure, each butterfly operation structure includes a multiplier, an adder and a subtractor,

[0011] The multiplier is used for calculation W The adder and subtractor are used to multiply the input b. a Add and subtract the output b*w of the multiplier to obtain the output values ​​A and B of the butterfly operation; where W represents the number of calculation points of the FFT unit.

[0012] The digital controlled oscillator module comprises a cascaded register step factor calculation unit, a phase register update unit and an interpolation parameter calculation unit, wherein the register step factor calculation unit is used for mapping the timing error to the register step factor parameter, the phase register update unit is used for updating the value of the phase register by the register step factor, and the interpolation parameter calculation unit determines the integer multiple interpolation sampling points and the decimal multiple interpolation sampling points by the value of the phase register;

[0013] The interpolation filter module includes a parallel branch filtering submodule and an interpolation filtering calculation unit, wherein each filtering unit in the parallel branch filtering submodule calculates the filtering branch result of the branch according to the integer multiple interpolation sampling points; the interpolation filtering calculation unit obtains the resampled signal after the timing synchronization is completed by weighted summing the output of the parallel branch filtering submodule and the decimal multiple interpolation sampling points; the parallel branch filtering submodule includes four filtering units arranged in parallel, each filtering unit also includes a filter parameter storage subunit, wherein the filter parameter storage subunit stores 4 groups of filtering parameters, and the filtering unit uses the filter parameters to perform weighted operations on the input signal and the input delayed signal respectively to obtain the filtering output of a single branch.

[0014] A timing synchronization method based on a chirp signal, using the chirp signal to calculate the timing error parameters caused by the timing drift, and using the peak position correction unit to correct the peak position of the spectrum channel during the FFT operation; after the timing error parameter estimation is completed, the resampling parameters of the received signal are calculated, and the received signal is resampled using an interpolation filter; the specific steps of the method are as follows:

[0015] Step 1, initializing the parameters of the receiver and transmitter;

[0016] Step 2, the timing error detection module calculates the timing error value of the chirped preamble signal based on the receiver and transmitter parameters initialized in step 1 to obtain a timing error parameter;

[0017] Step 3, the numerically controlled oscillator calculates the resampling parameters based on the timing error parameters obtained in step 2;

[0018] Step 4, the interpolation filter module obtains the timing synchronization result after the chirp signal is resampled in step 3;

[0019] Step 5: The system stops working.

[0020] The receiver and transmitter parameters in step 1 specifically include: receiving up-chirp signal, local de-chirp symbol d n And the coefficients c(p,q) of the cubic interpolation filter, the specific steps of initialization include:

[0021] Initialize the received up-chirp signal, including L up-chirp symbols x l , the Lth upper chirp symbol x l Includes N complex envelope signals c l (n), the nth complex envelope signal

[0022] Initialize the local de-chirp symbol to d n , whose length is N, the nth complex envelope signal and locally de-chirp the symbol d n stored in a de-chirp unit of a timing error detection module;

[0023] Initialize the coefficients c(p,q) of the cubic interpolation filter, and store the coefficients c(p,q) in the filter parameter storage subunit;

[0024] Wherein, δ represents the delay of an up-chirp symbol due to timing drift, L ≥ 10, N ≥ 512, -1 ≤ δ ≤ 1, p ∈ {1, 2, 3, 4}, q ∈ {-2, -1, 0, 1}, j represents the imaginary unit, and π represents the circumference of a circle.

[0025] The step 2 specifically includes:

[0026] Step 2.1: The de-chirping unit performs a chirp on each up-chirped symbol x. l The complex envelope signal c l With local de-chirp symbol d n Perform conjugate multiplication operation to obtain the de-chirped symbol de l , each de-chirped symbol l The expression is as follows:

[0027]

[0028] Step 2.2, the FFT operation unit performs the L de-chirped symbols de l The sequence is subjected to N-point FFT operation, and L groups of spectrum data F containing N elements are obtained. l ;

[0029] Step 2.3: The peak position detection unit detects L sets of spectrum data F l The maximum value of the modulus of the N elements is detected, and the L maximum modulus values ​​are obtained in F l The position w in lmax , and then take out the maximum value of each set of spectrum data and the frequency domain data F of the left and right elements respectively l (w lmax -1) F l (w lmax ), F l (w lmax+1);

[0030] Step 2.4, the peak position correction unit uses L sets of frequency domain data F l (w lmax -1) F l (w lmax ), F l (w lmax +1) Peak position correction factor Δw l Calculate and then use the peak position correction factor Δw l Calculate the peak correction positions of L groups of spectrum data respectively

[0031]

[0032] Step 2.5, the least squares fitting unit takes l as the independent variable and corrects the peak position of the spectrum data As the dependent variable, a line segment fitting is performed on the relationship curve between the upper chirp symbol number and the peak correction position of the spectrum data to obtain a linear equation y=δx+b, where:

[0033]

[0034] Step 2.6, the timing error parameter mapping unit performs parameter mapping on the timing error value Δ by fitting the slope δ of the curve. Since the delay variation in the N complex envelope signals of a chirp symbol is the slope δ of the fitting curve, the timing error value Δ in the scale of a single complex envelope signal is estimated:

[0035]

[0036] The step 3 specifically includes:

[0037] Step 3.1: The register step factor calculation unit calculates the step factor by using the timing error parameter Δ The calculation formula is as follows:

[0038]

[0039] Step 3.2, the phase register updates the unit by the step factor Updated step factor The value of the local phase register η n Update, where 2≤n≤N:

[0040]

[0041] Among them, (·) mod 1 means to perform a modulo operation on 1;

[0042] Step 3.3, the interpolation parameter calculation unit uses the value η of the local phase register n The changing zero-crossing condition, for integer multiples of interpolated sampling points m k And the fractional multiple interpolation sampling point u k Calculate: The value of the local phase register η n The elements in the Is it true? If so: set the current η n The serial number n is used as an integer multiple of the interpolation sampling point m k A sample of , and calculate the current decimal multiple interpolation sampling point Otherwise: the nth data is not used as an interpolation sampling point.

[0043] The step 4 specifically includes:

[0044] Step 4.1, the parallel branch filtering submodule reads out the coefficients c(p,q) of the local cubic interpolation filter and interpolates the sampling point m according to the integer multiples k The calculation results of the four parallel filter branches are given respectively:

[0045]

[0046] Through the above parallel branch filtering submodules, L up-chirp symbols x are obtained respectively. l The branch filter value y after parallel interpolation filtering lc1 ,y lc2 ,y lc3 ,y lc4 ;

[0047] Step 4.2, the interpolation filter calculation unit calculates the output y of the four parallel branches of the parallel branch filter submodule lc1 ,y lc2 ,y lc3 ,y lc4 The k elements of are interpolated and filtered in parallel, and the sampling points u are interpolated by fractional multiples k Perform weighted summation to obtain L up-chirp symbols x after timing synchronization is completed l The resampled signal

[0048]

[0049] Compared with the prior art, the beneficial effects of the present invention are:

[0050] First, the timing error detection module of the present invention includes a cascaded de-chirp signal peak position operation submodule and a timing error parameter calculation submodule. In the process of estimating the timing error parameters, compared with the traditional Gardner algorithm, the present invention converts the time domain timing error calculation to frequency domain processing, maps the timing error parameters to the maximum frequency spectrum position of each de-chirp symbol through the time-frequency correlation of the chirp signal, and solves the timing error parameters by least squares fitting the frequency spectrum position, effectively reducing the impact of noise on the estimation accuracy. This method can still maintain stable operation in a low signal-to-noise ratio environment, and significantly improves the noise resistance and robustness of timing synchronization compared to the prior art.

[0051] Secondly, the peak position correction unit in the de-chirp signal peak position calculation submodule of the present invention corrects the peak position of the spectrum result obtained by the FFT operation. Compared with the traditional FFT operation, this unit uses the spectrum maximum value position and its left and right spectrum data to correct the maximum value position, and then fits and maps the timing error signal through the spectrum channel position change parameters. Compared with the prior art, the estimation accuracy of the timing error value is improved.

[0052] Third, since the present invention calculates the timing error parameters in the frequency domain, it has a strong anti-noise performance. Therefore, compared with the Gardner algorithm, the loop filter module is omitted between the timing error detection module and the numerically controlled oscillator, which effectively reduces the hardware resource consumption compared with the prior art.

[0053] In summary, the present invention improves the anti-noise performance of timing synchronization while also improving the estimation accuracy of the timing error value. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a structural schematic diagram of the timing synchronization system of the present invention.

[0055] Figure 2 It is a structural schematic diagram of the timing error detection module of the present invention.

[0056] Figure 3 It is a structural schematic diagram of the FFT operation unit of the present invention.

[0057] Figure 4 It is a structural schematic diagram of the digital controlled oscillator module of the present invention.

[0058] Figure 5 It is a structural schematic diagram of the interpolation filter module of the present invention.

[0059] Figure 6 It is a structural schematic diagram of the filtering unit of the present invention.

[0060] Figure 7The figure is a flow chart of the timing synchronization method of the present invention.

[0061] Figure 8 This is a simulation diagram of the peak position correction unit of the simulation experiment of the present invention.

[0062] Fig. 9 It is a least square fitting unit simulation curve diagram of the simulation experiment of the present invention.

[0063] Fig.10 It is a simulation curve diagram of the residual MSE of the timing drift tracking loop of the simulation experiment of the present invention. DETAILED DESCRIPTION

[0064] The present invention will be described in detail below in conjunction with the accompanying drawings.

[0065] The technical idea for achieving the purpose of the present invention is: first, the delay of the received chirp symbol is estimated by the local de-chirp signal, and the delay parameter is estimated by the least squares fitting method, and the timing error parameter is obtained by parameter mapping. Then, the step factor of the local numerically controlled oscillator is calculated using the timing error parameter, and the phase register is updated according to the factor. Subsequently, based on the updated phase information, the integer multiple interpolation sampling points and the fractional multiple interpolation sampling points are calculated. Finally, the received signal is resampled by the interpolation filter to complete the timing synchronization of the signal.

[0066] See also Figure 1 , a chirp signal-based timing synchronization system, comprising a cascaded timing error detection module, a digitally controlled oscillator module and an interpolation filter module;

[0067] See also Figure 2The timing error detection module includes a cascaded de-chirp signal peak position operation submodule and a timing error parameter calculation submodule; the de-chirp signal peak position operation submodule includes a cascaded de-chirp unit, an FFT unit, a peak position detection unit, and a peak position correction unit, wherein the de-chirp unit conjugates and multiplies each up-chirp symbol with a local de-chirp symbol to obtain de-chirp symbols; the FFT unit performs spectrum operation on the de-chirp symbols to obtain spectrum data of each de-chirp symbol; the peak position detection unit detects the maximum value position of the spectrum data modulus value of each de-chirp symbol, and simultaneously extracts the maximum value at The peak position correction unit corrects the position of the maximum value of the frequency domain data of each de-chirped symbol through the maximum value spectrum data and the spectrum data on the left and right sides to obtain the peak correction position of the spectrum data; the timing error parameter calculation submodule includes a cascaded least square fitting unit and a timing error parameter mapping unit, wherein the least square fitting unit performs a line segment fitting on the relationship curve between the chirp symbol sequence number and the spectrum data peak correction position to obtain a timing error value of a chirp symbol; the timing error parameter mapping unit calculates the timing error value within the scale of a single complex envelope signal through the timing error value of a chirp symbol.

[0068] See also Figure 3 The FFT unit includes a plurality of cascaded butterfly operation structures, each butterfly operation structure includes parallel arrangement A butterfly operation structure is provided, each butterfly operation structure includes a multiplier, an adder and a subtractor, wherein the multiplier is used to calculate the product of W and the input b, the adder and the subtractor are used to add a to and subtract the output b*w of the multiplier from a respectively, to obtain the output values ​​A and B of the butterfly operation; wherein W represents the number of calculation points of the FFT unit.

[0069] See also Figure 4 The digital controlled oscillator module includes a cascaded register step factor calculation unit, a phase register update unit and an interpolation parameter calculation unit, wherein the register step factor calculation unit is used to map the timing error to the register step factor parameter, the phase register update unit is used to update the value of the phase register by the register step factor, and the interpolation parameter calculation unit determines the integer multiple interpolation sampling points and the decimal multiple interpolation sampling points by the value of the phase register;

[0070] See also Figure 5 , Figure 6The interpolation filter module includes a parallel branch filtering submodule and an interpolation filtering calculation unit, wherein each filtering unit in the parallel branch filtering submodule calculates the filtering branch result of the branch according to the integer multiple interpolation sampling points; the interpolation filtering calculation unit obtains the resampled signal after the timing synchronization is completed by weighted summing the output of the parallel branch filtering submodule and the decimal multiple interpolation sampling points; the parallel branch filtering submodule includes four filtering units arranged in parallel, each filtering unit also includes a filter parameter storage subunit, wherein the filter parameter storage subunit stores 4 groups of filtering parameters, and the filtering unit uses the filter parameters to perform weighted operations on the input signal and the input delayed signal respectively to obtain the filtering output of a single branch.

[0071] Reference Figure 7 The synchronization method of the present invention comprises the following steps:

[0072] Step 1, initialization parameters:

[0073] Initialize the received up-chirp signal, including L up-chirp symbols x l , the Lth upper chirp symbol x l Includes N complex envelope signals c l (n), the nth complex envelope signal Initialize the local de-chirp symbol to d n , whose length is N, the nth complex envelope signal And d n Stored in the de-chirp signal storage unit of the timing error detection module; Initialize the coefficient c(p,q) of the cubic interpolation filter and store the coefficient c(p,q) in the filter parameter storage subunit; Initialize the initial value η of the local phase register 1 =1; where δ represents the delay of an up-chirp symbol due to timing drift, L≥10, N≥512, and j represents an imaginary unit; in this example, L=10, N=1024, and δ=0.2;

[0074] Step 2: The timing error detection module calculates the timing error value of the up-chirp preamble signal:

[0075] Step 2.1: The de-chirping unit performs a chirp on each up-chirped symbol x. l The complex envelope signal c l With local de-chirp symbol d n Perform conjugate multiplication operation to obtain the de-chirped symbol de l , each de-chirped symbol l The expression is as follows:

[0076]

[0077] in,(·) * Denotes the conjugate operation. l For analysis, 10 de-chirped symbols are l Are single frequency signals, respectively for de l The peak value position of the spectrum channel obtained after the 1024-point FFT operation is linearly related to the chirp symbol number l, and the slope of the linear relationship is δ. Therefore, the 10 groups of de-chirp symbols de can be used in the following. l Perform FFT operation and fit the maximum position of the spectrum channel to obtain a time delay δ of the up-chirp symbol.

[0078] Step 2.2, the FFT operation unit performs the 10 de-chirped symbols de l The sequence is subjected to 1024-point FFT operation, and 10 sets of spectrum data F containing 1024 elements are obtained respectively. l .

[0079] Step 2.3: The peak position detection unit detects 10 sets of spectrum data F l The maximum value of the modulus of the 1024 elements is detected, and the 10 maximum modulus values ​​are obtained respectively in F l The position w in lmax , and then take out the maximum value of each set of spectrum data and the frequency domain data F of the left and right elements respectively l (w lmax -1) F l (w lmax ), F l (w lmax +1).

[0080] Step 2.4, the peak position correction unit uses 10 sets of frequency domain data F l (w lmax -1) F l (w lmax ), F l (w lmax +1) Peak position correction factor Δw l Calculate and then use the peak position correction factor Δw l Calculate and obtain 10 sets of spectrum data peak correction positions respectively

[0081]

[0082] Step 2.5, the least squares fitting unit takes l as the independent variable and As the dependent variable, a line segment fitting is performed on the relationship curve between the upper chirp symbol number and the peak correction position of the spectrum data to obtain a linear equation y=δx+b, where:

[0083]

[0084] Step 2.6, the timing error parameter mapping unit performs parameter mapping on the timing error value Δ by fitting the slope δ of the curve. Since the delay variation in the N complex envelope signals of a chirp symbol is δ, the timing error value Δ in the scale of a single complex envelope signal can be estimated:

[0085]

[0086] Step 3, the numerically controlled oscillator calculates the resampling parameters:

[0087] Step 3.1, the phase register updates the unit by the step factor Updated step factor The value of the local phase register η n Update, where 2≤n≤N:

[0088]

[0089] Step 3.2, the phase register updates the unit by the step factor The value of the local phase register η n , 2≤n≤N is updated:

[0090]

[0091] Among them, (·) mod 1 represents the modulo operation on 1.

[0092] Step 3.3, the interpolation parameter calculation unit is calculated by η n The changing zero-crossing condition, for integer multiples of interpolated sampling points m k And the fractional multiple interpolation sampling point u k Calculate: for η n The elements in the Is it true? If so: set the current η n The serial number n is used as an integer multiple of the interpolation sampling point m k A sample of , and calculate the current decimal multiple interpolation sampling point Otherwise: the nth data is not used as an interpolation sampling point.

[0093] Step 4: The interpolation filter module obtains the timing synchronization result after the up-chirp signal is resampled:

[0094] Step 4.1, the parallel branch filtering submodule reads out the coefficients c(p,q) of the local cubic interpolation filter and interpolates the sampling point m according to the integer multiples k The calculation results of the four parallel filter branches are given respectively:

[0095]

[0096] Through the above parallel branch filtering submodules, 10 up-chirp symbols x are obtained respectively. l The branch filter value y after parallel interpolation filtering lc1 ,y lc2 ,y lc3 ,y lc4 .

[0097] Step 4.2, the interpolation filter calculation unit calculates the output y of the four parallel branches of the parallel branch filter submodule lc1 ,y lc2 ,y lc3 ,y lc4 The k parallel interpolation filter structure interpolates the sampling points u by fractional multiples k Perform weighted summation to obtain 10 up-chirp symbols x after timing synchronization is completed l The resampled signal

[0098]

[0099] Simulation experiment

[0100] In order to verify the performance of the present invention, the following simulations are performed according to the instantiation parameters to illustrate the performance of the present invention, including the peak position correction unit precision correction simulation example, the least squares fitting unit example, and the timing error detection result MSE curve example;

[0101] like Figure 8 As shown, according to the simulation example corrected by the spectrum channel peak position correction unit, taking the first up-chirp symbol as an example, its peak position is an integer multiple of the FFT unit output position w lmax =741, the peak position frequency domain output F l (w lmax ) and the frequency domain values ​​F at the left and right points l (w lmax -1) F l (w lmax +1) Take it out and get F l (w lmax )=-915.86-1i*13257、F l (w lmax -1)=-30.29-1i*419.69、F l (w lmax +1)=29.62+1i*448, according to the three FFT output results, the peak position correction unit calculates the formula for Δw l Calculate and get Δw l=0.0327, then Similarly, the above operation is performed on each chirp symbol to obtain the corrected spectrum channel peak position, and finally the fractional multiple correlation peak position is obtained to show a linear growth trend. The corrected peak position is symbolically fitted to obtain the curve shown in Figure 9;

[0102] Through the least square fitting unit, the chirp symbol number is used as the x-axis, and each decimal multiple correlation peak position is used as the y-axis to obtain a linear slope of 0.20316. The slope is taken as a timing drift error parameter N*Δ=0.20316 within the time scale of the chirp symbol. In the same timing drift scenario, the estimated error parameter MSE is compared with the Gardner algorithm, and the following is obtained: Fig.10 The curve shown;

[0103] from Fig.10 It can be seen that in a high signal-to-noise ratio environment, when the signal-to-noise ratio range SNR>0dB, due to the addition of the spectrum channel peak position correction unit, the accuracy is improved to a certain extent compared with the Gardner algorithm, and the mean square error MSE of the timing drift estimation error parameter is maintained at 1×10 -5 The mean square error of the timing drift estimation error parameter of the Gardner algorithm in this environment is kept at 2×10 -5 In a low signal-to-noise ratio environment, when the signal-to-noise ratio range SNR<0dB, the MSE of the Gardner algorithm time domain estimation algorithm gradually increases with the decrease of the signal-to-noise ratio. At a -5dB signal-to-noise ratio, the MSE value has reached 8×10 -5 , which has a certain impact on the demodulation accuracy of the system. Relatively speaking, due to the characteristics of the chirp signal timing drift frequency domain estimation, its estimated noise resistance performance is stronger, and the MSE can still be maintained at 2×10 in a -10dB signal-to-noise ratio environment. -5 about.

[0104] Application prospects

[0105] With the rapid development of satellite communication systems, changes in satellite altitude and orbit have led to the problem of signal timing drift. This problem is more prominent in low-orbit satellites. Taking the Ka band under a 600km satellite orbit as an example, the timing drift under a 1s time scale can reach 0.022ms. At the same time, the signal quality in the satellite scene is poor, which brings great challenges to the timing synchronization algorithm of the communication system. In this context, the timing synchronization system and method based on chirp signals of the present invention have important application prospects. The system can accurately estimate and compensate for the timing drift in low-orbit satellite communications under low signal-to-noise ratio, especially to cope with large-scale timing drift caused by high-speed satellite movement and low-orbit satellite scenes. By using least squares fitting to estimate the chirp symbol delay, mapping the timing error parameters, and interpolation resampling technology based on updated phase information, the timing error can be efficiently eliminated to ensure the synchronization accuracy of the system in a high dynamic environment. This method has broad application value in low-orbit satellite communications, navigation positioning, remote sensing monitoring and other fields. It can effectively improve the robustness and accuracy of signal synchronization, ensure the quality and stability of data transmission in high-speed mobile platforms, and promote the reliable operation and large-scale deployment of low-orbit satellite networks.

Claims

1. A timing synchronization system based on chirp signals, characterized in that: It includes a cascaded timing error detection module, a digitally controlled oscillator module and an interpolation filter module; The timing error detection module is used to calculate the peak position of the spectrum channel of the de-chirped single tone signal, and calculate the timing error parameter through the peak position of the spectrum channel; The digital controlled oscillator module is used to map the timing error to the register step factor parameter, update the value of the phase register through the register step factor, and determine the integer multiple interpolation sampling points and the fractional multiple interpolation sampling points; The interpolation filter module resamples the received chirp symbols through integer multiple interpolation sampling points and decimal multiple interpolation sampling points to complete timing synchronization.

2. A timing synchronization system based on chirp signals according to claim 1, characterized in that: The timing error detection module comprises a cascaded de-chirp signal peak position calculation submodule and a timing error parameter calculation submodule; The de-chirp signal peak position calculation submodule comprises a cascaded de-chirp unit, an FFT unit, a peak position detection unit, and a peak position correction unit, wherein the de-chirp unit conjugates and multiplies each up-chirp symbol with a local de-chirp symbol to obtain de-chirp symbols; the FFT unit performs spectrum calculation on the de-chirp symbols to obtain spectrum data of each de-chirp symbol; the peak position detection unit detects the maximum value position of the spectrum data modulus of each de-chirp symbol, and simultaneously extracts the spectrum data on the left and right sides including the maximum value; the peak position correction unit corrects the position of the maximum value of the frequency domain data of each de-chirp symbol through the maximum value spectrum data and the spectrum data on the left and right sides to obtain the spectrum data peak correction position; The timing error parameter calculation submodule includes a cascaded least square fitting unit and a timing error parameter mapping unit, wherein the least square fitting unit performs a line segment fitting on the chirp symbol sequence number and the spectrum data peak correction position relationship curve to obtain a timing error value of a chirp symbol; the timing error parameter mapping unit calculates the timing error value within the scale of a single complex envelope signal through the timing error value of a chirp symbol.

3. A timing synchronization system based on chirp signals according to claim 2, characterized in that: The FFT unit includes a plurality of cascaded butterfly operation structures, each butterfly operation structure includes parallel arrangement A butterfly operation structure, each butterfly operation structure includes a multiplier, an adder and a subtractor. The multiplier is used to calculate W The adder and subtractor are used to multiply the input b. a Add and subtract the output b*w of the multiplier to obtain the output values ​​A and B of the butterfly operation; where W represents the number of calculation points of the FFT unit.

4. The timing synchronization system based on chirp signal according to claim 1, characterized in that: The digital controlled oscillator module comprises a cascaded register step factor calculation unit, a phase register update unit and an interpolation parameter calculation unit, wherein the register step factor calculation unit is used for mapping the timing error to the register step factor parameter, the phase register update unit is used for updating the value of the phase register through the register step factor, and the interpolation parameter calculation unit determines the integer multiple interpolation sampling points and the decimal multiple interpolation sampling points through the value of the phase register.

5. The timing synchronization system based on chirp signal according to claim 1, characterized in that: The interpolation filter module comprises a parallel branch filtering submodule and an interpolation filtering calculation unit, wherein each filtering unit in the parallel branch filtering submodule calculates the filtering branch result of the branch according to the integer multiple interpolation sampling points; The interpolation filter calculation unit obtains a resampled signal after timing synchronization is completed by weighted summing the output of the parallel branch filter submodule and the decimal multiple interpolation sampling point; the parallel branch filter submodule includes four filter units arranged in parallel, each filter unit also includes a filter parameter storage subunit, wherein the filter parameter storage subunit stores 4 groups of filter parameters, and the filter unit uses the filter parameters to perform weighted operations on the input signal and the input delayed signal respectively to obtain the filter output of a single branch.

6. A timing synchronization method based on a chirp signal according to the system of any one of claims 1 to 5, characterized in that: The chirp signal is used to calculate the timing error parameters caused by the timing drift, and the peak position of the spectrum channel is corrected by the peak position correction unit during the FFT operation. After the timing error parameter estimation is completed, the resampling parameters of the received signal are calculated, and the received signal is resampled using an interpolation filter. The specific steps of this method are as follows: Step 1, initialize the receiver and transmitter parameters; Step 2, the timing error detection module calculates the timing error value of the chirped preamble signal based on the receiver and transmitter parameters initialized in step 1 to obtain a timing error parameter; Step 3, the numerically controlled oscillator calculates the resampling parameters based on the timing error parameters obtained in step 2; Step 4, the interpolation filter module obtains the timing synchronization result after the chirp signal is resampled in step 3; Step 5: The system stops working.

7. The timing synchronization method based on chirp signal according to claim 6, characterized in that: The receiver and transmitter parameters in step 1 specifically include: receiving up-chirp signal, local de-chirp symbol d n And the coefficients c(p,q) of the cubic interpolation filter, the specific steps of initialization include: Initialize the received up-chirp signal, including L up-chirp symbols x l , the Lth upper chirp symbol x l Includes N complex envelope signals c l (n), the nth complex envelope signal Initialize the local de-chirp symbol to d n , whose length is N, the nth complex envelope signal and locally de-chirp the symbol d n stored in a de-chirp unit of a timing error detection module; Initialize the coefficients c(p,q) of the cubic interpolation filter, and store the coefficients c(p,q) in the filter parameter storage subunit; Wherein, δ represents the delay of an up-chirp symbol due to timing drift, L ≥ 10, N ≥ 512, -1 ≤ δ ≤ 1, p ∈ {1, 2, 3, 4}, q ∈ {-2, -1, 0, 1}, j represents the imaginary unit, and π represents the circumference of a circle.

8. The timing synchronization method based on chirp signal according to claim 6, characterized in that: The step 2 specifically includes: Step 2.1: The de-chirping unit performs a chirp on each up-chirped symbol x. l The complex envelope signal c l With local de-chirp symbol d n Perform conjugate multiplication operation to obtain the de-chirped symbol de l , each de-chirped symbol l The expression is as follows: Step 2.2, the FFT operation unit performs the L de-chirped symbols de l The sequence is subjected to N-point FFT operation, and L groups of spectrum data F containing N elements are obtained. l ; Step 2.3: The peak position detection unit detects L sets of spectrum data F l The maximum value of the modulus of the N elements is detected, and the L maximum modulus values ​​are obtained in F l The position w in lmax , and then take out the maximum value of each set of spectrum data and the frequency domain data F of the left and right elements respectively l (w lmax -1) F l (w lmax ), F l (w lmax +1); Step 2.4, the peak position correction unit uses L sets of frequency domain data F l (w lmax -1) F l (w lmax )、Fl(w lmax +1) Peak position correction factor Δw l Calculate and then use the peak position correction factor Δw l Calculate the peak correction positions of L groups of spectrum data respectively Step 2.5, the least squares fitting unit takes l as the independent variable and corrects the peak position of the spectrum data As the dependent variable, a line segment fitting is performed on the relationship curve between the upper chirp symbol number and the peak correction position of the spectrum data to obtain a linear equation y=δx+b, where: Step 2.6, the timing error parameter mapping unit performs parameter mapping on the timing error value Δ by fitting the slope δ of the curve. Since the delay variation in the N complex envelope signals of a chirp symbol is the slope δ of the fitting curve, the timing error value Δ in the scale of a single complex envelope signal is estimated:

9. The timing synchronization method based on chirp signal according to claim 6, characterized in that: The step 3 specifically includes: Step 3.1: The register step factor calculation unit calculates the step factor by using the timing error parameter Δ The calculation formula is as follows: Step 3.2, the phase register updates the unit by the step factor Updated step factor The value of the local phase register η n Update, where 2≤n≤N: Among them, (·) mod 1 means to perform a modulo operation on 1; Step 3.3, the interpolation parameter calculation unit uses the value η of the local phase register n The changing zero-crossing condition, for integer multiples of interpolated sampling points m k And the fractional multiple interpolation sampling point u k Calculate the value of the local phase register η n The elements in the Is it true? If so: set the current η n The serial number n is used as an integer multiple of the interpolation sampling point m k A sample of , and calculate the current decimal multiple interpolation sampling point Otherwise: the nth data is not used as an interpolation sampling point.

10. The timing synchronization method based on chirp signal according to claim 6, characterized in that: The step 4 specifically includes: Step 4.1, the parallel branch filtering submodule reads out the coefficients c(p,q) of the local cubic interpolation filter and interpolates the sampling point m according to the integer multiples k The calculation results of the four parallel filter branches are given respectively: Through the above parallel branch filtering submodules, L up-chirp symbols x are obtained respectively. l The branch filter value y after parallel interpolation filtering lc1 ,y lc2 ,y lc3 ,y lc4 ; Step 4.2, the interpolation filter calculation unit calculates the output y of the four parallel branches of the parallel branch filter submodule lc1 ,y lc2 ,y lc3 ,y lc4 The k elements of are interpolated and filtered in parallel, and the sampling points u are interpolated by fractional multiples k Perform weighted summation to obtain L up-chirp symbols x after timing synchronization is completed l The resampled signal

Images