FSK Carrier Frequency Offset Blind Estimation Method, System and Medium Based on L&R Algorithm

Abstract

The present invention discloses a blind estimation method, system and medium for the carrier frequency offset of FSK based on the L&R algorithm. The method includes the following steps: conjugately multiplying each local modulation sequence with the baseband signal of the FSK signal respectively to obtain corresponding first signals, the number of local modulation sequences corresponding to the modulation order of the FSK signal, and the local modulation sequences being different from each other; low-pass filtering and accumulating all the first signals to obtain a second signal with the modulation information eliminated; discretizing the second signal and then calculating the accumulated sum of the second signal, and substituting the accumulated sum of the second signal into the L&R algorithm to obtain the frequency offset estimation value of the FSK signal. The present invention can perform frequency offset estimation without prior information and has strong practicability.

Description

Technical Field

[0001] The present invention relates to the field of communication signal processing, and particularly to a blind estimation method, system and medium for FSK carrier frequency offset based on the L&R algorithm. Background Art

[0002] In a cooperative communication system, due to the inconsistency of physical hardware at the transceiver end or the relative movement between the two receiving parties, the signal carriers are out of sync. Carrier synchronization is a prerequisite for the normal operation of the system, especially in coherent receivers. At the same time, in communication detection and countermeasure, it is necessary to identify signals and obtain useful information. As an important parameter in a communication system, the signal carrier frequency, for example, most modulation signals require quadrature demodulation, so it is necessary to accurately estimate the carrier frequency, perform down-conversion according to the carrier frequency value, and shift the signal to the baseband, so as to better perform subsequent digital signal processing and obtain information from the signal; at the same time, in a coherent demodulation communication system, it is also necessary to accurately estimate the frequency offset of the signal in order to demodulate the received signal.

[0003] FSK modulation, as a common communication modulation method in communication systems, is widely used in various communication systems. Its demodulation algorithms are divided into coherent demodulation and non-coherent demodulation. Whether it is coherent demodulation or non-coherent demodulation, it is sensitive to frequency offset. Coherent demodulation is more sensitive to frequency offset, and its normalized mean square error of frequency offset estimation needs to be less than 10-4. Therefore, frequency offset estimation for an FSK modulation communication system is an essential step to improve system performance. The frequency offset estimation methods can be mainly divided into two categories: data-aided (DA) and non-data-aided (NDA). Most DA algorithms are based on maximum likelihood estimation, and classic algorithms among them include the Fitz algorithm, Kay algorithm, L&R algorithm, and M&M algorithm, etc. The L&R algorithm was proposed by Luise and Reggiannini in 1995 and is a fast, open-loop, all-digital algorithm. NDA algorithms are mainly improved based on the aforementioned DA algorithms.

[0004] The essence of the data-aided L&R algorithm is that it can accurately obtain and eliminate modulation information, thereby reducing estimation errors. For the non-data-aided L&R algorithm, currently, the receiving end needs to estimate the modulation information and then remove the modulation information from the received signal.

[0005] In most cases, the accuracy of DA algorithms is higher than that of NDA. However, since DA requires sending training sequences, it sacrifices part of the system communication efficiency. At the same time, DA algorithms require the receiver to know the sent training sequences, so they are mainly applied to cooperative communication. NDA algorithms do not require sending pilot signals, have high system bandwidth utilization, and at the same time, they do not require any prior information and can be used in non-cooperative communication scenarios. Therefore, applying the non-data-aided L&R algorithm to FSK modulation carrier frequency offset estimation has research value. Summary of the Invention

[0006] The technical problem to be solved by the present invention is: aiming at the technical problems existing in the prior art, the present invention provides an FSK carrier frequency offset blind estimation method, system and medium based on the L&R algorithm, which can perform frequency offset estimation without prior information and has strong practicability.

[0007] To solve the above technical problems, the technical solution proposed by the present invention is:

[0008] An FSK carrier frequency offset blind estimation method based on the L&R algorithm, comprising the following steps:

[0009] S1) Multiply each local modulation sequence by the conjugate of the baseband signal of the FSK signal respectively to obtain corresponding first signals. The number of local modulation sequences corresponds to the modulation order of the FSK signal, and the local modulation sequences are different from each other;

[0010] S2) Low-pass filter and accumulate all the first signals to obtain a second signal with modulation information eliminated;

[0011] S3) Discretize the second signal and calculate the sum of the second signal. Substitute the sum of the second signal into the L&R algorithm to calculate the frequency offset estimation value of the FSK signal.

[0012] Further, after step S3), it further includes: if the frequency offset estimation value of the FSK signal does not meet the accuracy requirement, perform frequency offset correction according to the frequency offset estimation value of the FSK signal and the baseband signal of the FSK signal to obtain a new baseband signal of the FSK signal, and execute the step of multiplying each local modulation sequence by the conjugate of the baseband signal of the FSK signal respectively.

[0013] Further, step S3) includes: calculating the current frequency offset estimation value according to the L&R algorithm and the second signal, updating the cumulative frequency offset estimation value according to the current frequency offset estimation value, and using the updated cumulative frequency offset estimation value as the frequency offset estimation value of the FSK signal.

[0014] Further, before step S1), it further includes the step of setting local modulation sequences according to the base system of the FSK signal, specifically including:

[0015] If the FSK signal is 2FSK, set the all-0 modulation sequence and the all-1 modulation sequence as local modulation sequences respectively;

[0016] If the FSK signal is 4FSK, set the all-0 modulation sequence to the all-3 modulation sequence as local modulation sequences respectively;

[0017] If the FSK signal is 8FSK, set the all-0 modulation sequence to the all-7 modulation sequence as local modulation sequences respectively.

[0018] Further, in step S2, low-pass filtering and accumulating all the first signals includes:

[0019] Passing each first signal through a corresponding low-pass filter respectively, and then adding the signals output by all the low-pass filters;

[0020] Or, adding all the first signals to obtain a third signal, and passing the third signal through a low-pass filter.

[0021] Further, before step S3, it further includes: accumulating each symbol sampling point in the second signal to obtain a new second signal.

[0022] Further, if the FSK signal is 2FSK and the second signal is e j2πΔft , the expression of the frequency offset estimation value of the FSK signal in step S3 is as follows:

[0023]

[0024] where R(m) = Me j(2πΔfkT+θ) , θ represents the signal initial phase, M represents the oversampling multiple, m represents the sampling moment, k represents the symbol sampling moment, T represents the symbol period, and L represents the number of symbols participating in averaging.

[0025] The present invention also provides an FSK carrier frequency offset blind estimation system based on the L&R algorithm, including:

[0026] A first calculation unit, configured to conjugate multiply each local modulation sequence with the baseband signal of the FSK signal respectively to obtain corresponding first signals, the number of local modulation sequences corresponding to the modulation order of the FSK signal, and the local modulation sequences being different from each other;

[0027] An accumulation and filtering unit, configured to low-pass filter and accumulate all the first signals to obtain a second signal with modulation information eliminated;

[0028] A second calculation unit, configured to calculate the frequency offset estimation value of the FSK signal according to the L&R algorithm and the second signal.

[0029] The present invention also provides an FSK carrier frequency offset blind estimation system based on the L&R algorithm, the computer being programmed or configured to execute any one of the FSK carrier frequency offset blind estimation methods based on the L&R algorithm.

[0030] The present invention also provides a computer-readable storage medium, in which a computer program programmed or configured to execute any one of the FSK carrier frequency offset blind estimation methods based on the L&R algorithm is stored.

[0031] Compared with the prior art, the advantages of the present invention are:

[0032] The present invention sets a corresponding number of local sequences based on the radix of the FSK signal, multiplies them conjugately with the baseband signal of the FSK signal respectively, adds these conjugately multiplied signals, and after low-pass filtering, the remaining frequency offset term after removing the modulation information is obtained. According to the L&R algorithm, the frequency offset estimation value is calculated from the remaining frequency offset term, so that the frequency offset estimation can be carried out without prior information. Compared with the traditional L&R algorithm, the algorithm complexity is not increased, and it has strong practicability.

[0033] The present invention can adjust the number of iterations of the algorithm according to the specific requirements for the frequency offset estimation accuracy. After calculating the frequency offset estimation value, the frequency offset correction is performed according to the current frequency offset estimation value and the baseband signal of the FSK signal, and then the corrected baseband is used to perform the frequency offset estimation again to obtain a new frequency offset estimation value until the index requirements are met, which has high flexibility. Brief Description of the Drawings

[0034] Figure 1 It is a flowchart of Embodiment 1 of the present invention.

[0035] Figure 2 It is a simplified flowchart of Embodiment 1 of the present invention.

[0036] Figure 3 It is a comparison diagram of the effects under 32 symbols in Embodiment 1 of the present invention.

[0037] Figure 4 It is a comparison diagram of the effects under 128 symbols in Embodiment 1 of the present invention.

[0038] Figure 5 It is a flowchart of Embodiment 2 of the present invention.

[0039] Figure 6 It is a comparison diagram of the effects in Embodiment 2 of the present invention. Detailed Embodiment

[0040] The present invention will be further described below in conjunction with the accompanying drawings of the specification and specific preferred embodiments, but the protection scope of the present invention is not limited thereby.

[0041] Embodiment 1

[0042] Before introducing the specific solution of this embodiment, the relevant concepts are introduced as follows:

[0043] Frequency offset estimation L&R algorithm

[0044] Assume that the signal used for carrier synchronization at the receiving end has completed symbol synchronization, and the inter-symbol interference is eliminated by channel equalization, and we can get

[0045] z(k) = w(k)exp[j(2πΔfkT + φ)] + n(k) k = 1, 2,... N (1)

[0046] where \(z(k)\) is the signal of carrier synchronization that has completed symbol synchronization and eliminated inter-symbol interference through channel equalization, \(\Delta f\) is the carrier frequency offset, \(\varphi\) is the initial phase, \(T\) is the symbol period, \(w(k)\) is the instantaneous sampling value of the received data, \(n(k)\) is independent additive Gaussian noise with a mean of zero and the real and imaginary parts having the same variance \(\sigma\). 2 , \(k\) is the sampling point corresponding to different sampling times, and \(N\) is the total number of sampling points.

[0047] The maximum likelihood estimation criterion is used to estimate the frequency offset \(\Delta f\), which is to make the conditional probability density function \(p(z|w(k),\Delta f,\varphi)\) reach the maximum. The likelihood probability density function is

[0048]

[0049] Simplifying the above formula gives

[0050]

[0051] where

[0052]

[0053] By transmitting a training sequence, the exact value of \(w(k)\) can be obtained in the receiver, and at this time, the data-aided method can be used to obtain the carrier parameters.

[0054] Since \(|z(k)-w(k)e\) j(2πΔfkT+φ) |\ 2 \(=|z(k)|\) 2 \(+|w(k)|\) 2 \(-2Re\{z(k)w\) * (k)e\) -j(2πΔfkT+φ)}], where the first two terms are independent of the parameter to be estimated, so the simplified likelihood function can be obtained

[0055]

[0056] Thus, the maximum likelihood estimation of the data-aided spectrum is

[0057]

[0058] The above formula can be equivalent to

[0059]

[0060] Let

[0061]

[0062] Then

[0063]

[0064] Let get

[0065]

[0066] Let \(x(k)=z(k)w\) * (k), then there is

[0067]

[0068] The autocorrelation function of \(x(k)\) in the data-aided algorithm is

[0069]

[0070] Since there is

[0071] \(x(k)=\rho(k)e\) j[2πΔfkT+θ+ψ(k)] (13)

[0072] where \(\theta\) is the initial phase, \(\psi(k)\) is the instantaneous phase at the \(k\) sampling point, and there is

[0073]

[0074] Therefore, it can be obtained that

[0075] \(R(m)=e\) j2πΔfkT +n”(m) \(0\leq m\leq N - 1\) (15)

[0076] Sum both sides of the above formula and take the average to smooth the noise term

[0077]

[0078] where \(L\) represents the number of symbols participating in the averaging. Ignoring the noise term, there is

[0079]

[0080] Examine the following equation

[0081]

[0082] When time, \(\sin\pi L\Delta fT / \sin\pi\Delta fT\) is a positive number, there is

[0083]

[0084] Thus, the frequency offset estimate value of \(\Delta f\) can be obtained

[0085]

[0086] where the frequency offset estimation range of the L&R algorithm is

[0087]

[0088] For signals without data assistance, the modulation information w(k) is unknown at the receiving end, and x(k) = z(k)w * (k) in the above text cannot be calculated. For non-data-assisted algorithms of PSK-like modulation signals, by averaging the likelihood probability density function p(z|w(k), Δf, φ) over the symbol w(k), we can obtain

[0089]

[0090] where M is the order of PSK modulation. Since the non-data-assisted algorithm for PSK modulation signals removes the modulation information through the operation of z M (k). However, for FSK-like modulation, the modulation information cannot be removed in this way, so it is not applicable to FSK-like modulation.

[0091] This embodiment proposes a blind estimation method for FSK carrier frequency offset based on the L&R algorithm, which removes the modulation information w(k) in FSK-like modulation signals and realizes the application of the non-data-assisted L&R algorithm to the blind estimation of the frequency offset of FSK signals. As Figure 1 shown, it includes the following steps:

[0092] S1) Multiply each local modulation sequence by the conjugate of the baseband signal of the FSK signal respectively to obtain the corresponding first signal. The number of local modulation sequences corresponds to the modulation order of the FSK signal, and the local modulation sequences are different from each other;

[0093] S2) Low-pass filter and accumulate all the first signals to obtain a second signal that eliminates the modulation information;

[0094] S3) Calculate the frequency offset estimation value of the FSK signal according to the L&R algorithm and the second signal.

[0095] Before step S1 of this embodiment, there is also a step of setting the local modulation sequence according to the base of the FSK signal, which specifically includes:

[0096] If the FSK signal is 2FSK, set 2 local modulation sequences, namely all-0 modulation sequence and all-1 modulation sequence;

[0097] If the FSK signal is 4FSK, set 4 local modulation sequences, namely all-0, 1, 2, 3 modulation sequences;

[0098] If the FSK signal is 8FSK, set 8 local modulation sequences, namely all-0, 1, 2, 3, 4, 5, 6, 7 modulation sequences.

[0099] In step S2 of this embodiment, low-pass filtering and accumulating all the first signals includes:

[0100] Each first signal is respectively passed through a corresponding low - pass filter, and then the signals output by all the low - pass filters are added together; to reduce the computational complexity, the process can also be simplified. For example, Figure 2 as shown, all the first signals are added together to obtain a third signal, and the third signal is passed through a low - pass filter. After simplification, the resources of the low - pass filter can be saved, but its performance remains unchanged.

[0101] In this embodiment, according to the order of the L&R algorithm, the frequency offset estimation range of the L&R algorithm can be calculated, and thus the lowest cut - off frequency of the low - pass filter can be obtained.

[0102] For example, Figure 1 and Figure 2 as shown, before step S3 of this embodiment, it also includes: accumulating each symbol sampling point in the second signal to obtain a new second signal. Since there is oversampling in the communication system, one symbol must correspond to multiple sampling points. Therefore, accumulating each symbol sampling point before step S3 can improve the system signal - to - noise ratio. The specific way of accumulating each symbol sampling point is to accumulate the sampling points within the target symbol to obtain the energy of the symbol.

[0103] Taking 2FSK as an example below, each step will be described.

[0104] In step S1, the all - 0 modulation sequence and the all - 1 modulation sequence are respectively conjugated - multiplied with the base - band IQ data to eliminate part of the modulation information. The derivation is as follows:

[0105] The base - band signal expression of 2FSK is

[0106]

[0107] where f d is the frequency interval of 2FSK, a n is the modulation information corresponding to the transmitted sequence, and its value is ±1, and Δf is the frequency offset value.

[0108] The local sequence expression corresponding to the all - 0 modulation sequence is:

[0109]

[0110] The local sequence expression corresponding to the all - 1 modulation sequence is:

[0111]

[0112] The conjugate multiplications of them with the base - band signal of 2FSK respectively can obtain the first signal as follows:

[0113]

[0114] Step S2 conjugates and multiplies the baseband signal of 2FSK with the all-0 modulation sequence and the all-1 modulation sequence respectively, and the corresponding first signals are low-pass filtered and summed to eliminate the influence of the modulation information. In Equation (26), when a n is 1, after the two formulas are low-pass filtered, the term passes through the filter, is filtered out by the low-pass filter. Similarly, when a n is -1, after the above two formulas in Equation (26) are low-pass filtered, the term passes through the filter, is filtered out by the low-pass filter. Therefore, the residual term after low-pass filtering, that is, the second signal, is always e j2πΔft , that is, the sum of the signals after low-pass output in Equation (26) is e j2πΔft . Discretizing it gives:

[0115]

[0116] where m is the sampling time, T s is the sampling period, n is the sampling time corresponding within one symbol, and its value range is from 0 to M - 1, M is the oversampling multiple, T is the symbol period, and k is the symbol sampling time. The value of the second signal accumulated within one symbol is:

[0117]

[0118] In the formula, θ is the initial phase of the signal, and its value is From (28) and Equation (15), it can be seen that e j(2πΔfkT+θ) is the value of R(m) without the noise term, where the initial phase is offset by θ. Therefore, using this signal to enter the L&R algorithm can estimate the frequency offset (since it is to find the phase value of R(m), the coefficient M has no influence on the calculation of the phase).

[0119] Step S3 substitutes the accumulated sum of the second signal, that is, R(m) = Me j(2πΔfkT+θ) , into Equation (20) in the derivation process of the L&R algorithm, and the expression of the frequency offset estimate value of the FSK signal can be obtained as follows:

[0120]

[0121] where R(m) = Me j(2πΔfkT+θ) , θ represents the initial phase of the signal, M represents the oversampling multiple, m represents the sampling time, k represents the symbol sampling time, T represents the symbol period, and L represents the number of symbols participating in the averaging.

[0122] Next, the effect of the method in this embodiment will be described.

[0123] Set the symbol rate of the 2FSK signal to 20 kHz, and the system sampling rate to 320 kHz. Simulate the method of this embodiment according to the number of symbols for estimating the carrier frequency being 32 and 128 respectively, and compare the simulation results with the data-aided L&R algorithm and the Cramer-Rao lower bound. The results are as Figure 3 and Figure 4 shown. In the figure, the line represents the simulation result of this embodiment. It can be seen from Figure 3 that the method of this embodiment can effectively estimate the frequency offset of the FSK modulated signal. At high signal-to-noise ratios, its estimation performance is close to the Cramer-Rao lower bound. It can be seen from Figure 4 that as the number of symbols used for frequency offset estimation decreases, the estimation accuracy of the method of this embodiment gradually improves.

[0124] In summary, this embodiment can estimate the system spectrum without prior information. Compared with the traditional L&R algorithm, it does not increase the algorithm complexity and has strong practicability.

[0125] Embodiment 2

[0126] This embodiment is basically the same as Embodiment 1, except that in order to improve the estimation accuracy, a multi-stage iterative method is used to improve the accuracy without changing the frequency offset estimation range.

[0127] As Figure 5 shown, after step S3) of this embodiment, it further includes: if the estimated value of the frequency offset of the FSK signal does not meet the accuracy requirement and the number of iterations has not been reached, perform frequency offset correction on the basis of the estimated value of the frequency offset of the FSK signal and the baseband signal of the FSK signal to obtain a new baseband signal of the FSK signal, and execute the step of multiplying each local modulation sequence by the conjugate of the baseband signal of the FSK signal respectively.

[0128] Correspondingly, step S3) of this embodiment includes: calculating the current frequency offset estimated value according to the L&R algorithm and the second signal, updating the cumulative frequency offset estimated value according to the current frequency offset estimated value, and using the updated cumulative frequency offset estimated value as the frequency offset estimated value of the FSK signal.

[0129] In this embodiment, the initial value of the cumulative frequency offset estimated value is 0. Therefore, the frequency offset estimated value of the FSK signal obtained in the first iteration is the frequency offset estimated value calculated in this iteration, and the frequency offset estimated value of the FSK signal obtained in subsequent iterations is the sum of all frequency offset estimated values up to this iteration.

[0130] Next, the effect of the method of this embodiment will be described.

[0131] Set the symbol rate of the 2FSK signal to 20 kHz, the system sampling rate to 320 kHz, and the number of iterations to 2. Among them, the number of symbols for the first carrier frequency estimation is 32, and the number of symbols for the second carrier frequency estimation is 128. Simulate the method of this embodiment and compare the simulation results with the data-aided L&R algorithm and the Cramer-Rao lower bound. The results are as Figure 6 shown. In the figure the line represents the simulation result of this embodiment. As Figure 6 can be seen, in the second iteration of the method of this embodiment, the estimation performance is consistent with the Cramer-Rao lower bound with 128 symbols for estimating the carrier frequency. Its spectrum estimation range can reach the estimation range of the L&R algorithm corresponding to 32 symbols for estimating the carrier frequency, and the performance is far better than that of the data-aided L&R frequency offset estimation algorithm with 32 symbols for estimating the carrier frequency.

[0132] In summary, this embodiment is flexible in the estimation range. It can adjust the number of iterations according to the specific requirements of the system for the frequency offset estimation accuracy to meet the index requirements, with high flexibility. At the same time, it does not require data-aided information, reducing the system overhead.

[0133] Embodiment 3

[0134] Based on Embodiment 1, this embodiment proposes an FSK carrier frequency offset blind estimation system based on the L&R algorithm, including:

[0135] A first calculation unit for conjugately multiplying each local modulation sequence with the baseband signal of the FSK signal respectively to obtain a corresponding first signal. The number of local modulation sequences corresponds to the modulation order of the FSK signal, and the local modulation sequences are different from each other;

[0136] An accumulation filtering unit for low-pass filtering and accumulating all the first signals to obtain a second signal that eliminates the modulation information;

[0137] A second calculation unit for calculating the frequency offset estimation value of the FSK signal according to the L&R algorithm and the second signal.

[0138] This embodiment also proposes an FSK carrier frequency offset blind estimation system based on the L&R algorithm. The computer is programmed or configured to execute the FSK carrier frequency offset blind estimation method based on the L&R algorithm described in Embodiment 1.

[0139] This embodiment also proposes a computer-readable storage medium, in which a computer program is stored that is programmed or configured to execute the FSK carrier frequency offset blind estimation method based on the L&R algorithm described in Embodiment 1.

[0140] The above are only the preferred embodiments of the present invention and do not impose any form of limitation on the present invention. Although the present invention has been disclosed above in its preferred embodiments, it is not intended to limit the present invention. Therefore, any simple modifications, equivalent changes, and decorations made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention shall fall within the scope of protection of the technical solution of the present invention.

Claims

1. A blind estimation method for FSK carrier frequency offset based on the L&R algorithm, characterized in that Including the following steps: S1) Respectively conjugate multiply each local modulation sequence with the baseband signal of the FSK signal to obtain corresponding first signals. The number of local modulation sequences corresponds to the modulation order of the FSK signal, and the local modulation sequences are different from each other; S2) Low-pass filter and accumulate all the first signals to obtain a second signal with the modulation information eliminated; S3) After discretizing the second signal, calculate the accumulated sum of the second signal, and substitute the accumulated sum of the second signal into the L&R algorithm to obtain the frequency offset estimation value of the FSK signal. The formula for the frequency offset estimation value is as follows: where R(m) is the cumulative sum of the second signal, m is the sampling time, L represents the number of symbols involved in averaging, and T is the symbol period.

2. The FSK carrier frequency offset blind estimation method based on the L&R algorithm according to claim 1, wherein After step S3), it further includes: If the frequency offset estimation value of the FSK signal does not meet the accuracy requirement, perform frequency offset correction based on the frequency offset estimation value of the FSK signal and the baseband signal of the FSK signal to obtain a new baseband signal of the FSK signal, and execute the step of respectively conjugate multiplying each local modulation sequence with the baseband signal of the FSK signal.

3. The FSK carrier frequency offset blind estimation method based on the L&R algorithm according to claim 2, wherein Step S3) includes: Calculate the current frequency offset estimation value according to the L&R algorithm and the second signal, update the accumulated frequency offset estimation value according to the current frequency offset estimation value, and use the updated accumulated frequency offset estimation value as the frequency offset estimation value of the FSK signal.

4. The FSK carrier frequency offset blind estimation method based on the L&R algorithm according to claim 1, wherein Before step S1), it further includes the step of setting local modulation sequences according to the base of the FSK signal, specifically including: If the FSK signal is 2FSK, set the all-0 modulation sequence and the all-1 modulation sequence as the local modulation sequences respectively; If the FSK signal is 4FSK, set the all-0 modulation sequence to the all-3 modulation sequence as the local modulation sequences respectively; If the FSK signal is 8FSK, set the all-0 modulation sequence to the all-7 modulation sequence as the local modulation sequences respectively.

5. The FSK carrier frequency offset blind estimation method based on the L&R algorithm according to claim 1, characterized in that In step S2), low-pass filtering and accumulating all the first signals includes: Respectively pass each first signal through the corresponding low-pass filter, and then add the signals output by all the low-pass filters; Or, add all the first signals to obtain a third signal, and pass the third signal through the low-pass filter.

6. The FSK carrier frequency offset blind estimation method based on the L&R algorithm according to claim 1, characterized in that Before step S3), it further includes: Accumulate the symbol sampling points in the second signal to obtain a new second signal.

7. The FSK carrier frequency offset blind estimation method based on the L&R algorithm according to claim 1, characterized in that, If the FSK signal is 2FSK and the second signal is e j2πΔft , in the expression of the frequency offset estimation value of the FSK signal in step S3, R(m) = Me j(2 πΔfkT+θ) , θ represents the signal initial phase, M represents the oversampling multiple, and k represents the symbol sampling time.

8. A blind estimation system for FSK carrier frequency offset based on the L&R algorithm, characterized in that, Including: A first calculation unit for respectively conjugate multiplying each local modulation sequence with the baseband signal of the FSK signal to obtain corresponding first signals. The number of local modulation sequences corresponds to the modulation order of the FSK signal, and the local modulation sequences are different from each other; An accumulation and filtering unit for low-pass filtering and accumulating all the first signals to obtain a second signal with the modulation information eliminated; A second calculation unit for calculating the frequency offset estimation value of the FSK signal according to the L&R algorithm and the second signal. The formula for the frequency offset estimation value is as follows: where R(m) is the cumulative sum of the second signal, m is the sampling time, L represents the number of symbols participating in averaging, and T is the symbol period.

9. An FSK carrier frequency offset blind estimation system based on the L&R algorithm, including a computer, characterized in that, The computer is programmed or configured to execute the FSK carrier frequency offset blind estimation method based on the L&R algorithm according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program programmed or configured to execute the FSK carrier frequency offset blind estimation method based on the L&R algorithm according to any one of claims 1 to 7.

Images