A method and apparatus for frequency offset estimation and compensation

Abstract

The present application relates to a kind of frequency offset estimation and compensation method and device, the method performs multi-stage frequency offset estimation, the method includes: after starting frequency offset estimation, according to the number of current execution frequency offset estimation stage to determine signal acquisition point number and fill the signal collected in buffer according to the signal acquisition point number;The filled signal in buffer is amplitude spectrum to obtain the maximum position of frequency offset, and when needed, the result of multiple complex modulus operation in amplitude spectrum process is accumulated in frequency offset estimation accumulator before the maximum position of frequency offset is obtained, and the maximum position of frequency offset is determined according to the result after accumulation operation;Frequency offset estimation value is calculated based on the maximum position of frequency offset, and frequency offset compensation is carried out according to the accumulation value of the frequency offset estimation value.Can realize wide range, high-precision frequency offset estimation and compensation.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of communication technology, in particular to a frequency offset estimation and compensation method and device. BACKGROUND

[0002] In high-speed optical fiber communication system, the useful signal is attenuated when passing through the filter, and the frequency offset will cause the phase rotation of the baseband signal. If there is a large frequency offset, the receiver will make a mistake when demapping or the decoder will make a mistake when soft demodulating, which seriously affects the quality of communication. Therefore, it is necessary to estimate and compensate the frequency offset of the received signal, so that the frequency offset of the signal after frequency offset compensation is small enough to be within the range that does not affect the performance of the receiver. The frequency offset will change due to the environmental temperature of the laser. Under normal circumstances, such change is slow and the amount of frequency offset change is more than two orders of magnitude smaller than the frequency offset itself, which is called temperature drift. When designing the frequency offset estimation algorithm, the influence of temperature drift or humidity drift should be considered in addition to the frequency offset caused by the laser.

[0003] In addition, with the popularization and commercialization of 400G optical transmission technology and communication system, digital coherent optical communication technology gradually replaces the traditional IMDD scheme and becomes the mainstream technology for future long-distance high-speed transmission, because it has better sensitivity, higher spectral efficiency and stronger damage resistance. Compared with the traditional IMDD, the digital coherent optical receiver uses the interference of the local oscillator light and the signal light to obtain all the optical field information of the signal such as complex amplitude, polarization and frequency, thereby supporting high-order modulation formats such as Quadrature Phase Shift Keying (QPSK) and 16-symbol Quadrature Amplitude Modulation (QAM). And through the Digital Signal Processor (DSP) technology, the electrical signal after coherent detection can be filtered and equalized in the electrical domain to compensate for the linear or even nonlinear damage experienced by the signal. At this time, the influence of the frequency offset of the transmitting and receiving laser during transmission can be estimated and compensated in the digital domain through the baseband signal in the digital domain.

[0004] At present, although the method of estimating the frequency offset in the baseband digital domain can obtain high estimation accuracy, it needs to consume more computing resources and computing time. SUMMARY

[0005] The embodiments of the present application provide a frequency offset estimation and compensation method and device, which realize wide-range and high-precision frequency offset estimation and compensation.

[0006] In one aspect, the embodiments of the present application provide a frequency offset estimation and compensation method, characterized in that the method performs multi-stage frequency offset estimation, and the method comprises:

[0007] After starting the frequency offset estimation, the number of signal collection points is determined according to the number of the current frequency offset estimation level, and the collected signals are filled into the buffer according to the number of the signal collection points;

[0008] The amplitude spectrum of the signals filled in the buffer is calculated to obtain the maximum frequency offset position, and if necessary, the results of the multiple modulus operations in the process of calculating the amplitude spectrum are accumulated in the frequency offset estimation accumulator before the maximum frequency offset position is obtained, and the maximum frequency offset position is determined according to the results after the accumulation operation;

[0009] The frequency offset estimation value is calculated based on the maximum frequency offset position, and the frequency offset compensation is performed according to the accumulated value of the frequency offset estimation value.

[0010] In some embodiments, the frequency offset compensation according to the accumulated value of the frequency offset estimation value includes the steps of:

[0011] The peak-to-average ratio of the data in the frequency offset estimation accumulator is calculated;

[0012] The validity of the currently calculated frequency offset estimation value is determined according to the peak-to-average ratio, if it is determined to be valid, the currently calculated frequency offset estimation value is accumulated into the frequency offset compensation register, and the frequency offset compensation value is determined according to the updated accumulated value in the frequency offset compensation register;

[0013] If it is determined to be invalid, the currently calculated frequency offset estimation value is discarded, and the frequency offset compensation value is determined according to the existing accumulated value in the frequency offset compensation register;

[0014] The frequency offset compensation is performed based on the frequency offset compensation value.

[0015] In some embodiments, the number of signal collection points is determined according to the number of the current frequency offset estimation level, and the collected signals are filled into the buffer according to the number of the signal collection points, including the steps of:

[0016] Setting the initial sampling point number P and the sampling frequency f s ;

[0017] According to P, wherein, is the number of the current frequency offset estimation level, is the preset sampling interval, and P is the initial sampling point number;

[0018] The received signal is sampled at the frequency f s , and the signal values of w sampling points are filled into the matching buffer.

[0019] In some embodiments, the amplitude spectrum of the signals filled in the buffer is calculated to obtain the maximum frequency offset position, including the steps of:

[0020] performing a fourth power or quadruple angle operation on the filled signals in the buffer as a complex vector X to obtain a complex vector Y;

[0021] performing an FFT operation on the complex vector Y to obtain a complex vector Z;

[0022] performing a complex modulus operation on the complex vector Z to obtain a real vector Z';

[0023] determining a maximum position A according to the real vector Z', and if A is greater than half of the number of signal collection points in the current frequency offset estimation, then A is subtracted by the number of signal collection points and the result is taken as the frequency offset maximum position A'.

[0024] In some embodiments, the results of the multiple complex modulus operations in the amplitude spectrum determination process are accumulated in a frequency offset estimation accumulator, and the frequency offset maximum position is determined according to the accumulated results, including the steps of:

[0025] setting a repetition number R;

[0026] performing R times of signal collection and filling, fourth power or quadruple angle operation, discrete Fourier transform operation, and complex modulus operation in the current frequency offset estimation, and accumulating the modulus values after the complex modulus operation in the frequency offset estimation accumulator;

[0027] determining a maximum position A according to the modulus value accumulation results in the frequency offset estimation accumulator, and if A is greater than half of the number of signal collection points in the current frequency offset estimation, then A is subtracted by the number of signal collection points and the result is taken as the frequency offset maximum position A'.

[0028] In some embodiments, the frequency offset estimation value is calculated based on the frequency offset maximum position, including the steps of:

[0029] calculating the frequency offset estimation value according to a first formula, the first formula including:

[0030] fo_est = 256 / * A',

[0031] wherein fo_est is the normalized frequency offset estimation value, N is the number of the current frequency offset estimation, and A' is the frequency offset maximum position.

[0032] In some embodiments, after the frequency offset estimation value is calculated based on the frequency offset maximum position, the method further includes:

[0033] determining whether the number of the current frequency offset estimation reaches a first preset threshold, and if the first preset threshold is not reached, then the number of the current frequency offset estimation is increased by one and the frequency offset estimation is restarted.

[0034] In some embodiments, the step of determining the validity of the frequency offset estimation value calculated according to the peak-to-average ratio when performing the current frequency offset estimation comprises the steps of:

[0035] If the peak-to-average ratio is not lower than a preset threshold, determining that the current calculated frequency offset estimation value is valid.

[0036] If the peak-to-average ratio is lower than the preset threshold, determining that the current calculated frequency offset estimation value is invalid.

[0037] In some embodiments, the method further comprises the steps of:

[0038] If the number of times of determining that the current calculated frequency offset estimation value is invalid reaches a second preset threshold, setting the number of the current frequency offset estimation level to 1, restarting the frequency offset estimation, and clearing the frequency offset compensation register.

[0039] In a second aspect, the embodiments of the present application further provide a frequency offset estimation and compensation device, characterized in that it comprises:

[0040] a signal collection module, configured to determine the number of signal collection points according to the number of the current frequency offset estimation level after starting the frequency offset estimation, and fill the collected signals into a buffer according to the number of signal collection points;

[0041] a frequency offset maximum position determination module, configured to calculate the amplitude spectrum of the signals filled in the buffer to obtain the frequency offset maximum position, and when necessary, accumulate the results of the multiple modulus operations in the process of calculating the amplitude spectrum in the frequency offset estimation accumulator before obtaining the frequency offset maximum position, and determine the frequency offset maximum position according to the results after the accumulation operation;

[0042] a frequency offset compensation module, configured to calculate the frequency offset estimation value based on the frequency offset maximum position, and perform the frequency offset compensation according to the accumulated value of the frequency offset estimation value.

[0043] The embodiments of the present application provide a frequency offset estimation and compensation method and device. The embodiments of the present application design the frequency offset estimation and compensation of the to-be-estimated signal into a frequency offset compensation mechanism with a feedback loop, use multiple frequency offset estimations to improve the timeliness of tracking the frequency offset changes, select different sampling point numbers according to each level to make the estimation range of the first level frequency offset estimation larger, and reduce the estimation range of each subsequent level frequency offset estimation while increasing the accuracy, and each level estimation is performed on the basis of the compensation of the previous level, thereby meeting the needs of ensuring the estimation range and accuracy at the same time. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiments description. Obviously, the drawings in the following description only show some of the embodiments of the present application, and other drawings can be obtained by those of ordinary skill in the art without any creative effort based on these drawings.

[0045] Figure 1 A flowchart of a frequency offset estimation and compensation method provided by the embodiment of the present application is shown in the figure.

[0046] Figure 2 An accumulation operation diagram in a frequency offset estimation accumulator provided by the embodiment of the present application is shown in the figure.

[0047] Figure 3 A flowchart of a frequency offset estimation and compensation method provided by the embodiment of the present application is shown in the figure.

[0048] Figure 4 A structure diagram of a frequency offset estimation and compensation device provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0049] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the following will describe the technical solutions in the embodiments of the present application clearly and completely with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without any creative effort belong to the protection scope of the present application.

[0050] As shown in the figure, the embodiment of the present application provides a frequency offset estimation and compensation method, which includes: Figure 1

[0051] S100: After starting the frequency offset estimation, determining the signal collection point number according to the number of the series of the frequency offset estimation currently performed and filling the collected signal into the buffer according to the signal collection point number;

[0052] S200: Calculating the amplitude spectrum of the signal filled in the buffer to obtain the maximum frequency offset position;

[0053] S300: When necessary, before obtaining the maximum frequency offset position, performing the accumulation operation on the results of the multiple complex modulus operation in the amplitude spectrum calculation process in the frequency offset estimation accumulator and determining the maximum frequency offset position according to the results after the accumulation operation;

[0054] S400: Calculating the frequency offset estimation value based on the maximum frequency offset position and performing the frequency offset estimation according to the accumulation value of the frequency offset estimation value.

[0055] ​It is to be noted that the embodiment performs multi-stage frequency offset estimation, and each time the frequency offset estimation is performed on the basis of the previous frequency offset compensation. The number of stages of the multi-stage frequency offset estimation is associated with the selected multiple when sampling the signal. For example, for the received signal with a frequency f s When sampling, if the 64-point FFT is selected for the signal spectrum line after four times of squaring, the frequency interval between each adjacent spectrum line is fs / 4 / 64. That is, the range of the frequency offset estimation is ±f s s / 8 / 4, and the accuracy of the second-stage frequency offset estimation is (fs / 4) / 64 / 4. The input of the third-stage frequency offset estimation is the signal compensated by the first-stage and the second-stage, and the third-stage frequency offset estimation is performed on the signal with a 16-fold (16-fold is a selected value) decimation, the estimation range is ±fs / 8 / 16, and the accuracy of the third-stage frequency offset estimation is (fs / 4) / 64 / 16. The input of the fourth-stage frequency offset estimation is the signal compensated by the first-stage, the second-stage, and the third-stage, and the fourth-stage frequency offset estimation is performed on the signal with a 64-fold (64-fold is a selected value) decimation, the estimation range is ±fs / 8 / 64, and the accuracy of the fourth-stage frequency offset estimation is (fs / 4) / 64 / 64. The input of the fifth-stage frequency offset estimation is the signal compensated by the first-stage to the fourth-stage, and the fifth-stage frequency offset estimation is performed on the signal with a 256-fold (256-fold is a selected value) decimation, the estimation range is ±fs / 8 / 256, and the accuracy of the fifth-stage frequency offset estimation is (fs / 4) / 64 / 256. It can be understood that the accuracy of the fifth-stage frequency offset estimation is (fs / 4) / 64 / 256, that is, the final residual frequency offset is (fs / 4) / 64 / 256. For example, when fs is 60 GHz, no matter whether the frequency offset is 3 GHz or 0 Hz or other values, as long as the signal after the fifth-stage frequency offset estimation and compensation has a frequency offset of (60e9 / 4) / 64 / 256 = 915527 Hz, that is, the signal after the final frequency offset compensation has a residual frequency offset of about 1 MHz. Therefore, the estimation range of the frequency offset is smaller and smaller, and the estimation accuracy is more and more accurate.

[0056] It is to be noted that the amplitude spectrum in S200 can be sequentially subjected to the four times of squaring or the four times of angle operation, the discrete Fourier transform operation, and the complex modulus operation.

[0057] It can be understood that, in order to achieve better noise reduction, the accumulation operation of S300 can be selectively used to determine the maximum frequency offset position. If the working environment is very ideal (in the case of very small noise), the accumulation operation of S300 is not performed, and the maximum frequency offset position can be directly found after the four times of squaring or the four times of angle operation, the discrete Fourier transform operation, and the complex modulus operation.

[0058] The embodiment of the present application can perform multi-stage frequency offset estimation, and each time the frequency offset estimation is performed on the basis of the previous frequency offset compensation. The frequency offset estimation and compensation of the signal to be estimated are designed as a frequency offset compensation mechanism with a feedback loop, and the multi-stage frequency offset estimation improves the timeliness of tracking the frequency offset change. At the same time, different sampling point numbers are selected for each stage, so that the first stage of frequency offset estimation has a larger estimation range, and the estimation range of each subsequent stage is reduced while the accuracy is increased. Each stage of estimation is performed on the basis of the compensation of the previous stage, thereby meeting the needs of ensuring the estimation range and accuracy.

[0059] Further, a first preset threshold can be set to limit the maximum number of stages of performing frequency offset estimation. After each frequency offset estimation is completed (after the frequency offset estimation value is calculated), it is determined whether the number of stages of performing frequency offset estimation currently reaches the first preset threshold. If the first preset threshold is not reached, the number of stages of performing frequency offset estimation is increased by one and the frequency offset estimation is restarted.

[0060] It can be understood that, for example, the first preset threshold can be set to 5, and each time one stage of frequency offset estimation is completed, the number of stages is increased by one, from 1 to 5. After five stages of frequency offset estimation and compensation, the system will normally run on the fifth stage of frequency offset estimation (based on the fact that the frequency offset changes very slowly).

[0061] In some embodiments, the frequency offset compensation according to the accumulated value of the frequency offset estimation value in S400 includes the following steps:

[0062] S410: Calculate the peak-to-average ratio of the data in the frequency offset estimation accumulator.

[0063] S420: Determine the validity of the currently calculated frequency offset estimation value according to the peak-to-average ratio. If it is determined to be valid, the currently calculated frequency offset estimation value is accumulated in the frequency offset compensation register, and the frequency offset compensation value is determined according to the updated accumulated value in the frequency offset compensation register. If it is determined to be invalid, the currently calculated frequency offset estimation value is discarded, and the frequency offset compensation value is determined according to the existing accumulated value in the frequency offset compensation register.

[0064] S430: Perform frequency offset compensation based on the frequency offset compensation value.

[0065] In some embodiments, the determination of the number of signal acquisition points according to the number of stages of performing frequency offset estimation in S100 and the filling of the acquired signal into the buffer according to the number of signal acquisition points include the following steps:

[0066] S110: Set the initial sampling point number P and the sampling frequency f s ;

[0067] S120: Determine the number of signal acquisition points w according to P, wherein, the number of the current frequency offset estimation, P is a preset sampling interval, and P is the initial number of sampling points;

[0068] S130: taking the signal with the frequency f s The received signal is sampled and the signal values of w sampling points are filled into the matching buffer.

[0069] It can be understood that when the current frequency offset estimation level is N, every number of sampling points to take one number until P numbers are taken. For example, set P to 64, and when 1 level of frequency offset estimation is currently performed, the input signal of the continuous 64 sampling points (the signal sampling point number w is 64 at this time) is stored into the buffer in turn. When 2 levels of frequency offset estimation are currently performed, 1 sampling point of the input signal every 4 sampling points needs to be taken and stored into the buffer, and 256 sampling points (the signal sampling point number w is 256 at this time) of the input can fill the buffer. When 3 levels of frequency offset estimation are currently performed, 1 sampling point of the input signal every 16 sampling points needs to be taken and stored into the buffer, and 1024 sampling points of the input can fill the buffer. When 4 levels of frequency offset estimation are currently performed, 1 sampling point of the input signal every 64 sampling points needs to be taken and stored into the buffer, and 4096 sampling points of the input can fill the buffer. When 5 levels of frequency offset estimation are currently performed, 1 sampling point of the input signal every 256 sampling points needs to be taken and stored into the buffer, and 16384 sampling points of the input can fill the buffer.

[0070] In some embodiments, the signal filled in the buffer in S200 is subjected to four times or four times angle operation, discrete Fourier transform operation and complex modulus operation in turn to obtain the maximum frequency offset position, comprising the steps of:

[0071] S210: taking the signal filled in the buffer as a complex vector X to obtain a complex vector Y through four times or four times angle operation;

[0072] S220: performing FFT operation on the complex vector Y to obtain a complex vector Z;

[0073] S230: performing complex modulus operation on the complex vector Z to obtain a real vector Z';

[0074] S240: determining the maximum position A according to the real vector Z', if A is greater than half of the signal sampling point number of the current frequency offset estimation, then A is subtracted from the signal sampling point number and the result after subtraction is taken as the maximum frequency offset position A'.

[0075] Preferably, in S210, the signal filled in the buffer is taken as a complex vector X to perform the fourth power or the four times angle operation to obtain a complex vector Y, and the four times angle operation is performed to obtain the complex vector Y for the 16QAM probability shaping system, and the fourth power operation is performed to obtain the complex vector Y for the QPSK, 8QAM and 16QAM. The 16QAM probability shaping system is to map the bit stream information to the 16 constellation points of the 16QAM in a certain way with unequal probability, and the unequal probability standard is that the probability of mapping to the inner circle of the constellation diagram is large, and the probability of mapping to the outer circle of the constellation diagram is small. The problem of the fourth power FFT modulus accumulation result not being a peak caused by too few outer circle symbols in the probability shaping PCS (probability shaping) system can be solved by using exp(j*angle(x)*4).

[0076] In some embodiments, in S300, the results of the multiple complex modulus operations are accumulated in a frequency offset estimation accumulator, and the maximum position of the frequency offset is determined according to the results after the accumulation, including the steps of:

[0077] S310: setting the repetition number R;

[0078] S320: performing R times of signal acquisition and filling, fourth power or four times angle operation, discrete Fourier transform operation, complex modulus operation, and sequentially accumulating the modulus values after the complex modulus operation in the frequency offset estimation accumulator during the current frequency offset estimation;

[0079] S330: determining the maximum position A according to the modulus value accumulation result in the frequency offset estimation accumulator, and if A is greater than half of the signal acquisition point number of the current frequency offset estimation, then A is subtracted from the signal acquisition point number and the result after the subtraction is taken as the maximum position A' of the frequency offset.

[0080] It should be noted that the repetition number R can be determined by the current working signal-to-noise ratio, and in general, the more the number of accumulations, the less the estimation result is affected by the Gaussian additive noise, and the more reliable the estimation result is.

[0081] In this embodiment, in order to reduce the influence of noise, the maximum spectral line position is calculated after R times of accumulation by the fourth power FFT and the complex modulus operation. When performing multi-stage frequency offset estimation on the signal to be estimated, the frequency offset estimation range of each stage is reduced by R times, but the precision is increased by R times. At the same time, using the accumulation operation can also reduce the FFT length, effectively saving the hardware cost.

[0082] In some embodiments, in S400, when the frequency offset estimation value is calculated based on the maximum position of the frequency offset, the frequency offset estimation value can be calculated according to the first formula, and the first formula includes:

[0083] fo_est = 256 / A', wherein, fo est is the normalized frequency offset estimation value, N is the current execution number of frequency offset estimation, and A' is the maximum position of the frequency offset.

[0084] In some embodiments, the validity of the frequency offset estimation value calculated in the current execution of frequency offset estimation is determined according to the peak-to-average ratio in S420, including the steps of:

[0085] S421: if the peak-to-average ratio is not lower than a preset threshold, determining that the current calculated frequency offset estimation value is valid;

[0086] S422: if the peak-to-average ratio is lower than the preset threshold, determining that the current calculated frequency offset estimation value is invalid.

[0087] The preset threshold value can be selected in the range of [1.5, +∞). If the preset threshold value is too small, it is not easy to correct errors, and if the preset threshold value is too large, it is not easy to update the frequency offset value to be compensated in the frequency offset compensation register. The preset threshold value can be preferably set to 3.5 (selected through a large number of simulation results).

[0088] In some embodiments, considering that if the system cannot be synchronized for a long time, the frequency offset estimation and compensation need to be reset, a threshold value (second preset threshold value) of invalid times can be set. If the number of invalid times of the current calculated frequency offset estimation value reaches the second preset threshold value, the execution number of the current frequency offset estimation is set to 1, and the frequency offset compensation register is reset.

[0089] In a specific embodiment, the initial number of sampling points is selected to be 64, i.e., the input signal with frequency offset needs to be filled into a buffer with a length of 64, and the data in the buffer is recorded as a complex vector X. The first preset threshold value is set to 5. If the current frequency offset estimation is the first frequency offset estimation, the input signal of 64 consecutive sampling points is stored in the buffer in turn. If the current frequency offset estimation is the second frequency offset estimation, 1 sampling point of the input signal needs to be taken every 4 sampling points and stored in the buffer. A total of 256 sampling points of the input can fill the buffer. If the current frequency offset estimation is the third frequency offset estimation, 1 sampling point of the input signal needs to be taken every 16 sampling points and stored in the buffer. A total of 1024 sampling points of the input can fill the buffer. If the current frequency offset estimation is the fourth frequency offset estimation, 1 sampling point of the input signal needs to be taken every 64 sampling points and stored in the buffer. A total of 4096 sampling points of the input can fill the buffer. If the current frequency offset estimation is the fifth frequency offset estimation, 1 sampling point of the input signal needs to be taken every 256 sampling points and stored in the buffer. A total of 16384 sampling points of the input can fill the buffer.

[0090] For example, Figure 2As shown, when the buffer is full, the complex vector X of length 64 in the buffer is subjected to four times power or four times angle operation, FFT operation, complex modulus operation and accumulation in turn. Specifically, the complex vector X of length 64 stored in the buffer is subjected to four times power or four times angle operation to obtain Y = X4 or Y = exp(j*angle(X)*4), wherein four times angle is used for 16QAM probability shaping mode, four times power is used for QPSK, 8QAM and 16QAM. The 64-point FFT operation is performed on Y to obtain Z = FFT(Y). The complex modulus is performed on Z, and Z' is a real vector of length 64. Z' and the accumulator of length 64 are added, and the result of the addition is used to update the value of the accumulator, and the initial value of the accumulator is 64 0 values. That is, steps S121 and S123 are repeated 32 times to obtain 32 real vectors Z' of length 64.

[0091] The number of repetitions is set to 32, and Z' is accumulated 32 times according to different input complex vectors X to obtain the real vector accumulator of length 64 updated 32 times, so that the influence of noise can be greatly reduced.

[0092] Wherein, the formula accumulator_new = accumulator_lasttime + Z' can be used for updating the accumulator each time, wherein accumulator_new is the updated accumulator value, and accumulator_lasttime is the accumulator value before updating. Each level of frequency offset estimation needs to complete 32 accumulations.

[0093] As shown in Figure 3 The maximum value of the accumulator accumulator is found to estimate the frequency offset. The specific method includes: finding the maximum value position max_p of the accumulator, first judging whether the maximum value position max_p is greater than 32, if the maximum value position max_p is greater than 32, making the maximum value position max_p' = the maximum value position max_p - 64, and after this processing, the final maximum value position max_p' takes a value between -31 and 32.

[0094] It is continuously judged whether the peak-to-average ratio of the data in the accumulator is greater than a threshold, if yes, it is considered that the current frequency offset estimation is effective, and the normalized frequency offset estimation value fo_est is calculated, wherein fo_est = 256 / (4 N-1) max_p', N is the number of current frequency offset estimation, which can be 1 to 5, fo_est is the normalized frequency offset estimation result, which is reflected as an angular frequency. If no, it is judged whether it is below the threshold for 3 times continuously (the number of times below the threshold reaches 3 for the first time), if yes, it is judged that the frequency offset estimation state is abnormal, fo_est is cleared and the frequency offset estimation is restarted from the first stage. If it is not below the threshold for 3 times continuously (the number of times below the threshold does not reach 3), it is considered that the current frequency offset estimation is invalid (for example, the signal quality is very poor at this time, which is not enough for the frequency offset estimation module to work normally), and fo_est is set to zero.

[0095] According to the judgment result of whether the current frequency offset estimation is valid, it is determined whether to calculate the normalized frequency offset estimation value fo_est or to set fo_est to zero, and finally the result (calculated fo_est or 0) is output to the frequency offset compensation module CAR (i.e. the frequency offset compensation register for accumulation), and the accumulated value is used for frequency offset compensation. The initial value of the frequency offset compensation module CAR can be set to 0.

[0096] After judging that the current frequency offset estimation is valid, it is further judged whether the number of current frequency offset estimation is 5, if yes, the number of frequency offset estimation is kept unchanged at 5 and the accumulator is cleared and the frequency offset estimation is restarted. If it does not reach 5, the number of frequency offset estimation is increased by 1 and the accumulator is cleared and the frequency offset estimation is restarted.

[0097] Further, in order to enhance the robustness of the frequency offset estimation algorithm and avoid abnormal data impact, the frequency offset estimation runs to an abnormal state and does not return. In the judgment of whether the frequency offset estimation is valid, the peak-to-average ratio PAR is calculated according to PAR = max(accumulator) / average(accumulator), wherein max(accumulator) is the maximum value in the frequency offset estimation accumulator, and average(accumulator) is the average value in the frequency offset estimation accumulator.

[0098] On the other hand, as shown in FIG. 2, the embodiment of the present application also provides a frequency offset estimation and compensation device, which comprises: Figure 4 A signal acquisition module is used to determine the number of signal acquisition points according to the number of current frequency offset estimation after starting the frequency offset estimation, and fill the acquired signal into the buffer according to the number of signal acquisition points;

[0099]

[0100] ​The maximum frequency offset location determination module is used to calculate the amplitude spectrum of the signal filled in the buffer to obtain the maximum frequency offset location. When necessary, before obtaining the maximum frequency offset location, the results of multiple complex modulus operations during the amplitude spectrum calculation process are accumulated in the frequency offset estimation accumulator, and the maximum frequency offset location is determined based on the result of the accumulation operation.

[0101] The frequency offset compensation module is used to calculate the frequency offset estimate based on the maximum frequency offset position and perform frequency offset compensation based on the cumulative value of the frequency offset estimate.

[0102] In some embodiments, the frequency offset compensation module is also used for:

[0103] Calculate the peak-to-average power ratio (PAPR) of the data in the frequency offset estimation accumulator;

[0104] The validity of the currently calculated frequency offset estimate is determined based on the peak-to-average power ratio. If the estimate is valid, the currently calculated frequency offset estimate is added to the frequency offset compensation register, and the frequency offset compensation value is determined based on the updated accumulated value in the frequency offset compensation register.

[0105] If the result is deemed invalid, the currently calculated frequency offset estimate is discarded and the frequency offset compensation value is determined based on the existing accumulated value in the frequency offset compensation register.

[0106] Frequency offset compensation is performed based on the frequency offset compensation value.

[0107] In some embodiments, the signal acquisition module is also used for:

[0108] Set the initial number of sampling points P and the sampling frequency f. s ;

[0109] according to * P determines the number of signal acquisition points w, where, For the current frequency offset estimation level, The preset sampling interval is P, where P is the initial number of sampling points.

[0110] With frequency f s The received signal is sampled and the signal values ​​of w sample points are filled into the matching buffer.

[0111] In some embodiments, the frequency offset maximum location determination module is also used for:

[0112] The signal filled in the buffer is used as a complex vector X and subjected to a fourth power or a quadruple angle operation to obtain a complex vector Y.

[0113] Perform an FFT operation on the complex vector Y to obtain the complex vector Z;

[0114] Performing a complex modulus operation on the complex vector Z yields a real vector Z';

[0115] According to the real vector Z', a maximum position A is determined, if A is greater than half of the number of signal collection points in current frequency offset estimation, A is subtracted by the number of signal collection points and the result is taken as the frequency offset maximum position A'.

[0116] Further, the frequency offset maximum position determination module is further configured to:

[0117] set a repetition number R;

[0118] perform R times of signal collection and filling, fourth power or quadruple angle operation, discrete Fourier transform operation, complex number modulus operation and successively accumulate the modulus value after the complex number modulus operation in the frequency offset estimation accumulator;

[0119] According to the modulus value accumulation result in the frequency offset estimation accumulator, a maximum position A is determined, if A is greater than half of the number of signal collection points in current frequency offset estimation, A is subtracted by the number of signal collection points and the result is taken as the frequency offset maximum position A'.

[0120] Further, the frequency offset maximum position determination module calculates the frequency offset estimation value according to a first formula, the first formula comprising:

[0121] fo_est = 256 / * A',

[0122] wherein fo_est is the normalized frequency offset estimation value, N is the number of current frequency offset estimation, and A' is the frequency offset maximum position.

[0123] In some embodiments, after the frequency offset maximum position determination module calculates the frequency offset estimation value based on the frequency offset maximum position, it further judges whether the number of current frequency offset estimation reaches a first preset threshold, if the first preset threshold is not reached, the number of current frequency offset estimation is increased by one and the frequency offset estimation is restarted.

[0124] In some embodiments, when the frequency offset compensation module judges the validity of the frequency offset estimation value calculated in current frequency offset estimation according to the peak-to-average ratio, if the peak-to-average ratio is not lower than a preset threshold, it is judged that the current calculated frequency offset estimation value is valid; if the peak-to-average ratio is lower than the preset threshold, it is judged that the current calculated frequency offset estimation value is invalid.

[0125] In some embodiments, the frequency offset compensation module is further configured to, when the number of times of judging that the current calculated frequency offset estimation value is invalid reaches a second preset threshold, set the number of current frequency offset estimation to 1 and restart the frequency offset estimation, and clear the frequency offset compensation register.

[0126] Those skilled in the art can understand that all or some of the steps in the method disclosed above, the functions of the modules / units in the system and the device can be implemented by software, firmware, hardware, or a combination thereof. In a hardware implementation, the division between the functional modules / units referred to in the above description does not necessarily correspond to the division of physical components; for example, one physical component can have multiple functions, or one function or step can be performed by several physical components in cooperation. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on computer-readable storage media, which can include computer-readable storage media (or non-transitory media) and communication media (or transitory media).

[0127] It should be noted that the terms "first" and "second" and the like in the present disclosure are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between such entities or operations. Also, the terms "comprises", "comprising", or any other variations thereof are intended to cover a non-exclusive inclusion, so that a process, method, article, or apparatus that comprises a list of elements does not include only those elements recited, but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0128] The above description is merely one specific implementation of the application, and thus the skilled in the art can understand or implement the application without departing from the spirit or scope of the application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Therefore, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of frequency offset estimation and compensation, characterized by, The method performs multi-stage frequency offset estimation, and the method comprises: After starting the frequency offset estimation, the number of signal collection points is determined according to the number of stages of the current frequency offset estimation, and the collected signals are filled into the buffer according to the number of signal collection points; The filled signals in the buffer are subjected to amplitude spectrum to obtain a maximum frequency offset position, and when the noise is large, the results of multiple complex modulus operations in the process of amplitude spectrum are accumulated in a frequency offset estimation accumulator before the maximum frequency offset position is obtained, and the maximum frequency offset position is determined according to the results after the accumulation operation; The frequency offset estimation value is calculated based on the maximum frequency offset position, and the frequency offset compensation is performed according to the accumulated value of the frequency offset estimation value; It is judged whether the number of stages of the current frequency offset estimation reaches a first preset threshold, and if not, the number of stages of the current frequency offset estimation is increased by one and the frequency offset estimation is restarted; The number of signal collection points is determined according to the number of stages of the current frequency offset estimation, and the collected signals are filled into the buffer according to the number of signal collection points, which comprises the following steps: Set initial number of sampling points P and sampling frequency f s ; According to determining the signal collection point number w, wherein, for the current execution frequency offset estimation stage, is a preset sampling interval, and P is the initial sampling point number; at a frequency f s sampling the received signal and populating the matched buffer with signal values for w sample points; Each time the spectrum estimation is performed on the basis of the previous spectrum compensation.

2. The method of frequency offset estimation and compensation according to claim 1, wherein, The frequency offset compensation is performed according to the accumulated value of the frequency offset estimation value, which comprises the following steps: The peak-to-average ratio of the data in the frequency offset estimation accumulator is calculated; The validity of the currently calculated frequency offset estimation value is determined according to the peak-to-average ratio, if it is determined to be valid, the currently calculated frequency offset estimation value is accumulated in a frequency offset compensation register, and the frequency offset compensation value is determined according to the updated accumulated value in the frequency offset compensation register; If it is determined to be invalid, the currently calculated frequency offset estimation value is discarded, and the frequency offset compensation value is determined according to the existing accumulated value in the frequency offset compensation register; The frequency offset compensation is performed based on the frequency offset compensation value.

3. The frequency offset estimation and compensation method of claim 1, wherein, The filled signals in the buffer are subjected to amplitude spectrum to obtain a maximum frequency offset position, which comprises the following steps: The filled signals in the buffer are subjected to four times or four times angle operation to obtain a complex vector Y; The complex vector Y is subjected to FFT operation to obtain a complex vector Z; The complex vector Z is subjected to complex modulus operation to obtain a real vector Z'; The maximum position A is determined according to the real vector Z', if A is greater than half of the number of signal collection points in the current frequency offset estimation, A is subtracted by the number of signal collection points, and the result after the subtraction is taken as the maximum frequency offset position A'.

4. The method for frequency offset estimation and compensation according to claim 1, wherein, The results of multiple complex modulus operations in the process of amplitude spectrum are accumulated in the frequency offset estimation accumulator, and the maximum frequency offset position is determined according to the results after the accumulation operation, which comprises the following steps: The number of repetitions R is set; When the current frequency offset estimation is performed, the signal collection and filling, the four times or four times angle operation, the discrete Fourier transform operation, and the complex modulus operation are performed R times, and the modulus values after the complex modulus operation are accumulated in the frequency offset estimation accumulator; The maximum position A is determined according to the modulus value accumulation result in the frequency offset estimation accumulator, if A is greater than half of the number of signal collection points in the current frequency offset estimation, A is subtracted by the number of signal collection points, and the result after the subtraction is taken as the maximum frequency offset position A'.

5. The method for frequency offset estimation and compensation according to claim 1, wherein, The frequency offset estimation value is calculated based on the maximum frequency offset position, which comprises the following steps: The frequency offset estimation value is calculated according to a first formula, and the first formula comprises: ， Wherein, fo_est is a normalized frequency offset estimation value, N is a number of current frequency offset estimation, and A' is the maximum frequency offset position.

6. The method for frequency offset estimation and compensation according to claim 2, wherein, The validity of the frequency offset estimation value calculated when the current frequency offset estimation is performed is determined according to the peak-to-average ratio, and the method comprises the steps of: If the peak-to-average ratio is not lower than a preset threshold, it is determined that the current calculated frequency offset estimation value is valid. If the peak-to-average ratio is lower than the preset threshold, it is determined that the current calculated frequency offset estimation value is invalid.

7. The method for frequency offset estimation and compensation according to claim 1, wherein, The method further comprises the steps of: If the number of times when the current calculated frequency offset estimation value is determined to be invalid reaches a second preset threshold, the number of current frequency offset estimation is set to 1, and the frequency offset estimation is restarted, and the frequency offset compensation register is cleared.

8. A frequency offset estimation and compensation apparatus, characterized by comprising: The method comprises: A signal acquisition module is configured to determine a number of signal acquisition points according to the number of current frequency offset estimation after the frequency offset estimation is started, and fill the acquired signal into a buffer according to the number of signal acquisition points; A maximum frequency offset position determination module is configured to obtain the maximum frequency offset position by calculating the amplitude spectrum of the signal filled in the buffer, and when the noise is large, the results of multiple complex modulus operations in the process of calculating the amplitude spectrum are accumulated in a frequency offset estimation accumulator before the maximum frequency offset position is obtained, and the maximum frequency offset position is determined according to the results after the accumulation operation; A frequency offset compensation module is configured to calculate a frequency offset estimation value based on the maximum frequency offset position and perform frequency offset compensation according to the accumulated value of the frequency offset estimation value; After the maximum frequency offset position determination module calculates the frequency offset estimation value based on the maximum frequency offset position, it is further determined whether the number of current frequency offset estimation reaches a first preset threshold, and if the first preset threshold is not reached, the number of current frequency offset estimation is increased by one and the frequency offset estimation is restarted. The signal acquisition module is further configured to: Set initial number of sampling points P and sampling frequency f s ; According to determining the signal collection point number w, wherein, for the current execution frequency offset estimation stage, is a preset sampling interval, and P is the initial sampling point number; at a frequency f s sampling the received signal and populating the matched buffer with signal values for w sample points; Wherein, each time the frequency spectrum estimation is performed on the basis of the previous frequency spectrum compensation.

Images