A method for frequency offset estimation and correction in a severely multipath scattering channel

Abstract

The present invention discloses a frequency offset estimation and correction method in a severely multipath scattering channel, including: receiving a data frame that has passed through a multipath channel; respectively performing cross-correlation between a local CAZAC sequence, a local pilot sequence and the CAZAC sequence and the pilot sequence in the data frame to obtain a first channel estimation value and a second channel estimation value for each path in the multipath channel; respectively determining the weight of each path based on the first channel estimation value; performing frequency sweeping at a preset frequency offset interval according to a preset frequency offset range to obtain a plurality of different frequency offset values for frequency sweeping; screening out a frequency offset value for frequency sweeping from the plurality of different frequency offset values for frequency sweeping based on the weight and the first channel estimation value, and using the selected frequency offset value for frequency sweeping to correct the frequency offset of the data included in the data frame to obtain data after coarse frequency offset correction; and continuing to correct the frequency offset of the data after coarse frequency offset correction according to the second channel estimation value to obtain corrected data. The present invention can reliably perform frequency offset estimation in more multipaths and at a lower signal-to-noise ratio.

Description

Technical Field

[0001] The present invention belongs to the technical field of communication measurement and control, and particularly relates to a frequency offset estimation and correction method under a severely multipath scattering channel. Background Art

[0002] Due to the long single-hop span, anti-destruction, anti-interference, and anti-interception capabilities of scatter communication, and the characteristics of all-weather reliable communication across complex terrains such as mountains, bays, and deserts, it has always been highly regarded in military communication. Scatter communication has become an important means of modern military communication. In the field of communication countermeasures in modern electronic warfare, due to its advantages of good low intercept, anti-multipath, anti-near-far effect, and easy synchronization, spread spectrum communication technology plays an extremely important role. Spread spectrum communication exchanges the signal bandwidth for the benefit of signal-to-noise ratio, enabling information to be transmitted at a very low signal-to-noise ratio, thereby reducing the possibility of information being intercepted, and it is an important low probability of intercept technology. Currently, spread spectrum communication is generally used in satellite and military fields. Due to the relative high-speed movement between the mobile station and the base station, Doppler frequency offset will occur, and the baseband signal after coherent demodulation at the receiving end will still have a signal with this frequency offset, resulting in bit errors at the receiving end. The scatter channel is a typical multipath fading channel with severe frequency selective fading, and traditional frequency offset estimation methods cannot meet the frequency offset estimation requirements under multipath conditions. Therefore, solving the problem of large frequency offset estimation error when the aircraft moves in a low signal-to-noise ratio scatter multipath fading channel is a very important issue.

[0003] Frequency offset estimation is generally divided into time-domain frequency offset estimation methods and frequency-domain estimation methods. Among them, the frequency-domain estimation method is based on DFT and can be implemented through FFT. The position of the maximum spectral line of FFT corresponds to the frequency of the signal to be estimated. The time-domain estimation method is to add two CAZAC sequences of the same length at both ends of the data frame. At the receiving end, the two sequences of the same length are subjected to delayed cross-correlation to obtain the estimated frequency offset value. Usually, the choice of which estimation method can be determined by communication requirements.

[0004] However, the existing technologies have the following problems and defects:

[0005] (1) The existing frequency offset algorithms only perform frequency offset estimation in channels with single-digit multipaths. In the case of more multipaths in the scatter channel, the frequency offset estimation value is not accurate and differs greatly from the actual value.

[0006] (2) There is a fence effect in FFT in the frequency-domain estimation algorithm. Its frequency-domain resolution is the ratio of the sampling frequency to the number of FFT points. Only when the frequency to be estimated is an integer multiple of the frequency-domain resolution, the estimation will be accurate. Secondly, the existing PMT-FFT algorithm can only estimate large frequency offsets, and the estimation value is not very accurate either.

[0007] (3) In existing frequency offset estimation methods, traditional methods cannot reliably complete frequency offset estimation and correction at lower signal-to-noise ratios. Summary of the Invention

[0008] To solve the above problems existing in the related art, the present invention provides a frequency offset estimation and correction method in a severely multipath scattering channel. The technical problems to be solved by the present invention are achieved through the following technical solutions:

[0009] The present invention provides a frequency offset estimation and correction method in a severely multipath scattering channel, including:

[0010] Receiving a data frame that has passed through a multipath channel;

[0011] Performing cross-correlation between a local CAZAC sequence and the CAZAC sequence and pilot sequence included in the data frame respectively to obtain a first channel estimation value and a second channel estimation value for each path in the multipath channel

[0012] Determining the weight of each path based on the first channel estimation value;

[0013] Scanning frequencies at a preset frequency offset interval according to a preset frequency offset range to obtain a plurality of different scanned frequency offsets;

[0014] Based on the weight and the first channel estimation value, screening out a scanned frequency offset from the plurality of different scanned frequency offsets, and using the selected scanned frequency offset to correct the frequency offset of the data included in the data frame to obtain data after rough frequency offset correction;

[0015] Continuing to correct the frequency offset of the data after rough frequency offset correction according to the second channel estimation value to obtain corrected data.

[0016] In some embodiments, the data frame includes: two CAZAC sequences, two pilot sequences, and data; the performing cross-correlation between a local CAZAC sequence and the CAZAC sequence and pilot sequence included in the data frame respectively to obtain a first channel estimation value and a second channel estimation value for each path in the multipath channel includes:

[0017] Performing cross-correlation between two local CAZAC sequences and the two CAZAC sequences correspondingly to obtain a first channel estimation value for each path in the multipath channel;

[0018] Performing cross-correlation between the two local CAZAC sequences and the two pilot sequences correspondingly to obtain a second channel estimation value for each path in the multipath channel.

[0019] In some embodiments, the first channel estimation value includes a first sub-channel estimation value and a second sub-channel estimation value; the expression of the weight of each path is: where k = 0, …, N−1, N is the total number of paths, C1(k) is the first sub-channel estimation value of the k-th path, C2(k) is the second sub-channel estimation value of the k-th path, |·| represents the absolute value, and ∑ is the summation symbol.

[0020] In some embodiments, the first channel estimation value of each path includes: a first sub-channel estimation value and a second sub-channel estimation value; screening out one sweep frequency offset from the multiple different sweep frequency offsets based on the weight and the first channel estimation value includes:

[0021] Calculating the angular frequency corresponding to each sweep frequency offset to obtain a plurality of different angular frequencies in one-to-one correspondence with the multiple different sweep frequency offsets;

[0022] Calculating the corrected first sub-channel estimation value and the corrected second sub-channel estimation value of each path at each angular frequency according to the first sub-channel estimation value and the second sub-channel estimation value of each path;

[0023] Calculating the accumulation of the absolute value of the difference between the corrected first sub-channel estimation value and the corrected second sub-channel estimation value at each angular frequency to obtain the accumulated difference corresponding to each angular frequency;

[0024] Calculating the quantized accumulated difference corresponding to each angular frequency according to the accumulated difference corresponding to each angular frequency and the weight of each path;

[0025] Taking the sweep frequency offset corresponding to the angular frequency corresponding to the smallest quantized accumulated difference as the screened sweep frequency offset.

[0026] In some embodiments, the expression of the quantized accumulated difference corresponding to each angular frequency is:

[0027]

[0028] where J i is the quantized accumulated difference corresponding to the i-th angular frequency, i = 0, …, I−1, I is the total number of the multiple different angular frequencies, k = 0, …, N−1, N is the total number of paths, W k is the weight of the k-th path, C'1 k is the corrected first sub-channel estimation value of the k-th path, C'2 k is the corrected second sub-channel estimation value of the k-th path.

[0029] In some embodiments, the second channel estimation value includes: a third sub-channel estimation value and a fourth sub-channel estimation value; and the step of further performing frequency offset correction on the data after coarse frequency offset correction according to the second channel estimation value to obtain corrected data includes:

[0030] Calculating an average value between the third sub-channel estimation value and the fourth sub-channel estimation value included in the second channel estimation value of a preset path in the multipath channel;

[0031] Multiplying the conjugate of the average value by the data after coarse frequency offset correction to obtain the corrected data.

[0032] In some embodiments, the expression of the first sub-channel estimation value of each path is:

[0033]

[0034] The expression of the second sub-channel estimation value of each path is:

[0035]

[0036] where k = 0, …, N - 1, N is the total number of paths, n = 0, …, L, L + 1, …, 2L - 1, 2L is the total length of the local CAZAC sequence, p(n) is the local CAZAC sequence, p * (n) is the conjugate sequence of the local CAZAC sequence, f d is the frequency offset to be estimated, ω is the angular frequency corresponding to f d and ω = 2πf d .

[0037] In some embodiments, the expression of the corrected first sub-channel estimation value of each path at each swept frequency offset is:

[0038]

[0039] The expression of the corrected second sub-channel estimation value of each path at each swept frequency offset is:

[0040]

[0041] where k = 0, …, N - 1, N is the total number of the paths, n = 0, …, L, L + 1, …, 2L - 1, 2L is the total length of the local CAZAC sequence, ω i is the i-th angular frequency, i = 0, …, I - 1, I is the total number of the multiple different angular frequencies.

[0042] The present invention has the following beneficial technical effects:

[0043] By performing cross-correlation between the local CAZAC sequence and the CAZAC sequence and pilot sequence in the received data frame respectively, the first channel estimate value and the second channel estimate value of each path of the multipath channel are obtained. According to the first channel estimate value, the weight of each path is determined respectively. After roughly estimating the frequency offset from multiple known different swept frequency offsets based on the first channel estimate value and the weight, the data in the data frame is roughly corrected for the frequency offset by using the roughly estimated frequency offset. Finally, the data after the rough frequency offset correction is further finely corrected according to the second channel estimate value of a specific path of the multipath channel, so as to obtain the finally corrected data. In this way, the frequency offset estimation of the received signal can be combined with the maximum likelihood estimation algorithm and the multipath weighted combining technology, and the frequency offset of the received signal can be more accurately estimated in a multipath severe scattering channel and at a low signal-to-noise ratio, so that the frequency offset can be reliably estimated under more multipaths and lower signal-to-noise ratios.

[0044] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a flowchart of a frequency offset estimation and correction method in a multipath severe scattering channel provided by an embodiment of the present invention;

[0046] Figure 2 It is an exemplary pilot-data-pilot data frame structure provided by an embodiment of the present invention;

[0047] Figure 3A It is a schematic diagram of a PQSK symbol transmitted by a transmitter provided by an embodiment of the present invention;

[0048] Figure 3B It is a schematic diagram of a signal with a 15 Hz frequency offset received by a receiver provided by an embodiment of the present invention;

[0049] Figure 3C It is a schematic diagram of a signal after frequency offset correction provided by an embodiment of the present invention;

[0050] Figure 4 It is an MSE curve of an ML-based frequency sweeping method under different frequency offsets provided by an embodiment of the present invention;

[0051] Figure 5 It is an MSE comparison curve between an ML-based frequency sweeping method and a delayed autocorrelation method when the frequency offset is 35 Hz provided by an embodiment of the present invention;

[0052] Figure 6 It is an MSE comparison curve between an ML-based frequency sweeping method and a delayed autocorrelation method when the frequency offset is 5 Hz provided by an embodiment of the present invention

[0053] Figure 7 The MSE comparison curve of the frequency offset of -35 Hz for the ML-based frequency sweeping method and the delayed autocorrelation method provided by the embodiments of the present invention. Detailed implementation manners

[0054] The present invention will be further described in detail below in conjunction with specific embodiments, but the implementation manners of the present invention are not limited thereto.

[0055] In the description of the present invention, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present invention, "a plurality of" means two or more unless otherwise specifically defined.

[0056] In the description of this specification, the description with reference to terms such as "an embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic descriptions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, those skilled in the art can combine and combine the different embodiments or examples described in this specification.

[0057] Although the present invention has been described herein in connection with various embodiments, however, in the process of implementing the claimed invention, those skilled in the art can understand and achieve other variations of the disclosed embodiments by viewing the accompanying drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "one" does not exclude a plurality of cases. A single processor or other unit can implement several functions recited in the claims. Certain measures are recited in mutually different dependent claims, but this does not mean that these measures cannot be combined to produce good results.

[0058] Aiming at the problem that due to the severe multipath scattering channel in the existing algorithm, the traditional frequency offset estimation is inaccurate due to multipath fading, the present invention can more reliably complete the frequency offset estimation relative to the traditional algorithm in the severe multipath scattering channel through the provided frequency offset estimation and correction method; at the same time, a weighting coefficient is introduced for multipath merging, and the frequency offset estimation can be completed at a lower signal-to-noise ratio relative to the traditional frequency offset estimation algorithm.

[0059] Figure 1 It is a flowchart of the frequency offset estimation and correction method provided by an embodiment of the present invention in a severely multipath scattering channel. As Figure 1 shown, the method includes the following steps:

[0060] S101. Receive a data frame that has passed through a multipath channel.

[0061] Here, the channel impulse response of the multipath channel (scattering channel) can be expressed as h(n'), where n' = 0,..., N - 1, and N is the total number of paths in the multipath of the scattering channel.

[0062] Here, the data frame includes two CAZAC sequences, two pilot sequences, and data. Among them, the structures of the two pilot sequences and the data are as Figure 2 shown.

[0063] Here, the received data frame can be expressed as: n = 0,..., L, L + 1,..., 2L - 1, 2L is the total length of the CAZAC sequence sent by the transmitting end, ω is the angular frequency corresponding to the frequency offset f d and ω = 2πf d .

[0064] S102. Respectively perform cross-correlation between the local CAZAC sequence, the local pilot sequence and the CAZAC sequence and the pilot sequence included in the data frame to obtain the first channel estimate value and the second channel estimate value of each path in the multipath channel.

[0065] Here, two local CAZAC sequences can be used to perform cross-correlation with the two CAZAC sequences correspondingly to obtain the first channel estimate value of each path in the multipath channel; and two local pilot sequences can be used to perform cross-correlation with the two pilot sequences correspondingly to obtain the second channel estimate value of each path in the multipath channel.

[0066] Here, the local pilot sequence is composed of the local CAZAC sequence.

[0067] Here, the local CAZAC sequence can be expressed as p(n), where n = 0,..., L, L + 1,..., 2L - 1, and 2L is the total length of the local CAZAC sequence. Specifically, the local CAZAC sequence includes a first local CAZAC sequence and a second local CAZAC sequence with the same length. Correspondingly, the first channel estimate value of each path includes a first sub-channel estimate value C1(k) and a second sub-channel estimate value C2(k). The local pilot sequence includes a first local pilot sequence and a second local pilot sequence with the same length. Correspondingly, the second channel estimate value of each path includes a third sub-channel estimate value R1(k) and a fourth sub-channel estimate value R2(k).

[0068] Specifically, the first local CAZAC sequence can be cross-correlated with one of the CAZAC sequences in the data frame to obtain the first sub-channel estimation value of each path in the multipath of the received channel; the second local CAZAC sequence can be cross-correlated with another CAZAC sequence in the data frame to obtain the second sub-channel estimation value of each path.

[0069] Exemplarily, the calculation formula for the first sub-channel estimation value C1(k) of each path is:

[0070]

[0071] where p1(n) is the first local CAZAC sequence, and p1 * (n) is the conjugate sequence of p1(n).

[0072] Exemplarily, the calculation formula for the second sub-channel estimation value C2(k) of each path is:

[0073] where k = 0, …, N - 1, N is the total number of paths, n = 0, …, L, L + 1, …, 2L - 1, 2L is the total length of the local CAZAC sequence, p2(n) is the second local CAZAC sequence, and p2 * (n) is the conjugate sequence of p2(n), f d is the frequency offset to be estimated, ω is the angular frequency corresponding to f d and ω = 2πf d .

[0074] Here, the original expression of C1(k) is: The original expression of C2(k) is: Due to the characteristics of the CAZAC sequence, the second term in the original expressions of C1(k) and C2(k) can be ignored compared to the first term. Therefore, the above expressions of C1(k) and C2(k) can be obtained.

[0075] Specifically, the first local pilot sequence can be cross-correlated with one of the pilot sequences in the data frame to obtain the third sub-channel estimation value of each path in the multipath of the received channel; the second local pilot sequence can be cross-correlated with another pilot sequence in the data frame to obtain the fourth sub-channel estimation value of each path.

[0076] Exemplarily, the calculation formula for the third sub-channel estimation value R1(k) of each path is:

[0077]

[0078] where q1(n) is the first local pilot sequence, and q1* (n) is the conjugate sequence of q1(n).

[0079] Exemplarily, the calculation formula for the fourth sub-channel estimation value R2(k) of each path is:

[0080]

[0081] where q2(n) is the second local pilot sequence, and q2 * (n) is the conjugate sequence of q2(n).

[0082] S103. Determine the weight of each path based on the first channel estimation value respectively.

[0083] Specifically, the expression of the weight of each path is: |·| represents the absolute value, and ∑ is the summation symbol.

[0084] S104. Perform frequency sweeping according to a preset frequency offset range at a preset frequency offset interval to obtain multiple different frequency sweeping offsets.

[0085] Here, the preset frequency offset range can be arbitrarily set according to actual needs, and the preset frequency offset interval can also be set according to actual needs. The embodiments of the present invention do not limit this.

[0086] Exemplarily, I different frequency sweeping offsets can be obtained.

[0087] S105. Screen out a frequency sweeping offset from multiple different frequency sweeping offsets based on the weight and the first channel estimation value, and use the selected frequency sweeping offset to correct the frequency offset of the data included in the data frame to obtain the data after coarse frequency offset correction.

[0088] Here, the angular frequency corresponding to each frequency sweeping offset can be calculated to obtain I different angular frequencies corresponding one-to-one to the I different frequency sweeping offsets; then, according to the first sub-channel estimation value C1(k) and the second sub-channel estimation value C2(k) of each path k, calculate the corrected first sub-channel estimation value C'1(k) and the corrected second sub-channel estimation value C'2(k) of each path at each angular frequency; then, calculate the accumulation of the absolute value of the difference between the corrected first sub-channel estimation value C'1(k) and the corrected second sub-channel estimation value C'2(k) at each angular frequency ω i to obtain the cumulative difference J corresponding to each angular frequency i ; according to the cumulative difference J corresponding to each angular frequency i and the weight W of each path k , calculate the quantized cumulative difference corresponding to each angular frequency; take the frequency sweeping offset corresponding to the angular frequency corresponding to the smallest quantized cumulative difference as the selected frequency sweeping offset.

[0089] Exemplarily, each angular frequency ω i The expression of the corresponding quantized cumulative difference is: C'1 k is the corrected first sub-channel estimate of the k-th path, and C'2 k is the corrected second sub-channel estimate of the k-th path.

[0090] S106. According to the second channel estimate, continue to perform frequency offset correction on the data after coarse frequency offset correction to obtain the corrected data.

[0091] Here, the average value between the third sub-channel estimate R1(0) and the fourth sub-channel estimate R2(0) included in the second channel estimate of a preset path (for example, the path corresponding to k = 0) in the multipath channel can be calculated Multiply the conjugate R * (0) of the average value by the data after coarse frequency offset correction to obtain the corrected data.

[0092] Exemplarily, the expression of the corrected data is: S'(n) = S(n) × R * (0); where S(n) is the data after coarse frequency offset correction, and S'(n) is the corrected data.

[0093] The present invention overcomes the problem that traditional algorithms cannot reliably perform frequency offset estimation in severely multipath scattering channels and at low signal-to-noise ratios. By combining the maximum likelihood estimation algorithm and the multipath weighted merging technique, an ML-based time-domain frequency offset estimation algorithm is formed, which can more accurately perform frequency offset estimation on the received signal in severely multipath scattering channels and at low signal-to-noise ratios, so that frequency offset estimation can be reliably performed under more multipaths and lower signal-to-noise ratios.

[0094] The following further illustrates the technical effects that can be achieved by the method provided by the present invention through simulation experiment data.

[0095] In this chapter, MATLAB is used as the simulation platform to perform simulation analysis on the ML-based time-domain frequency offset estimation method (i.e., the frequency offset estimation and correction method provided by the present invention). The classification algorithms mainly include three types: the delayed autocorrelation method, the FFT frequency domain estimation method, and the ML-based time-domain frequency offset estimation method. In the simulation experiment, the bandwidth is 500 kHz, the signal-to-noise ratio is -24 dB to -10 dB, the transmitted signal is QPSK symbols, the frame length is 54 ms, the channel is an actually collected scattering channel with 120 multipaths, the frequency offset values are ±35 Hz, ±30 Hz, ±15 Hz, ±5 Hz, the length of each segment of the frequency offset estimation sequence is 9200, that is, the length of each segment is 4600. The data frame structure is a pilot-data-pilot structure, and the data volume ratio is 1:6:1. The frequency sweep interval during frequency offset estimation is 2 Hz. The number of simulation times is 3000 times.

[0096] Figure 3A is the PQSK symbol sent by the transmitter; Figure 3B is the signal with a 15 Hz frequency offset received by the receiver; Figure 3C is the signal after frequency offset correction.

[0097] Figure 4 are the MSE curves of the ML-based frequency sweep method under different frequency offsets, where the scanned frequency offsets are ±35 Hz, ±30 Hz, ±15 Hz, ±5 Hz. From Figure 4 it can be seen that when the signal-to-noise ratio is -24 dB and the frequency offset is within -30 Hz to 30 Hz, the MSE value is less than 100, that is, the coarse frequency offset estimation error is within 10 Hz. For frequency offsets above 30 Hz, the MSE value can drop to 100 when the signal-to-noise ratio is -23 dB, and at this time the coarse frequency offset estimation error is within 10 Hz.

[0098] Figure 5 is the MSE comparison curve between the ML-based frequency sweep method and the delayed autocorrelation method when the frequency offset is 35 Hz. Figure 6 is the MSE comparison curve between the ML-based frequency sweep method and the delayed autocorrelation method when the frequency offset is 5 Hz. Figure 7 is the MSE comparison curve between the ML-based frequency sweep method and the delayed autocorrelation method when the frequency offset is -35 Hz. From the above three curve graphs, it can be seen that for the delayed autocorrelation algorithm and the ML-based frequency sweep algorithm, the smaller the frequency offset, the more accurate the algorithm's estimation of the frequency offset. In a severely multipath scattering channel, the performance of the ML-based frequency sweep algorithm is about 8 dB better than that of the delayed autocorrelation algorithm.

[0099] The conclusion obtained from the above simulation is that in a severely multipath scattering channel, compared with the traditional delayed autocorrelation algorithm, the performance of the ML-based time-domain frequency sweep algorithm is improved by about 8 dB, that is, it can obtain a more accurate frequency offset estimation value at a lower signal-to-noise ratio.

[0100] The above content is a further detailed description of the present invention in combination with specific preferred embodiments. It cannot be determined that the specific implementation of the present invention is only limited to these descriptions. For those of ordinary skill in the technical field to which the present invention belongs, without departing from the concept of the present invention, several simple deductions or substitutions can still be made, which should all be regarded as belonging to the protection scope of the present invention.

Claims

1. A frequency offset estimation and correction method under a severely multipath scattering channel, characterized in that Comprising: Receiving a data frame that has passed through a multipath channel; The data frame includes: two CAZAC sequences, two pilot sequences, and data; Performing cross-correlation between two local CAZAC sequences and the two CAZAC sequences correspondingly to obtain a first channel estimate value for each path in the multipath channel; The first channel estimate value for each path includes: a first sub-channel estimate value and a second sub-channel estimate value; Performing cross-correlation between two local pilot sequences and the two pilot sequences correspondingly to obtain a second channel estimate value for each path in the multipath channel; The second channel estimate value includes: a third sub-channel estimate value and a fourth sub-channel estimate value; Based on the first channel estimate value, determining the weight of each path respectively; Scanning frequencies according to a preset frequency offset range at a preset frequency offset interval to obtain a plurality of different scanned frequency offsets; Calculating the angular frequency corresponding to each scanned frequency offset to obtain a plurality of different angular frequencies that are in one-to-one correspondence with the plurality of different scanned frequency offsets; Calculate the corrected first sub-channel estimate and the corrected second sub-channel estimate for each path at each angular frequency according to the first sub-channel estimate and the second sub-channel estimate of each path; the expression for the corrected first sub-channel estimate of each path at each swept frequency offset is: The expression for the corrected second sub-channel estimate of each path at each swept frequency offset is: where k = 0, …, N-1, N is the total number of the paths, n = 0, ..., L, L+1, ..., 2L-1, 2L is the total length of the local CAZAC sequence, ω i is the i-th angular frequency, i = 0, …, I-1, I is the total number of the multiple different angular frequencies; Calculating the cumulative sum of the absolute values of the differences between the corrected first sub-channel estimate value and the corrected second sub-channel estimate value at each angular frequency to obtain the cumulative difference corresponding to each angular frequency; Calculate the quantized cumulative difference corresponding to each angular frequency according to the cumulative difference corresponding to each angular frequency and the weight of each path; the expression of the quantized cumulative difference corresponding to each angular frequency is: where J i is the quantized cumulative difference corresponding to the i-th angular frequency, W k is the weight of the k-th path, C'1 k is the corrected first sub-channel estimate value of the k-th path, C'2 k is the corrected second sub-channel estimate value of the k-th path; Taking the scanned frequency offset corresponding to the angular frequency with the smallest quantized cumulative difference as the selected scanned frequency offset, and using the selected scanned frequency offset to correct the frequency offset of the data included in the data frame to obtain the data after coarse frequency offset correction; Calculating the average value between the third sub-channel estimate value and the fourth sub-channel estimate value included in the second channel estimate value of a preset path in the multipath channel; Multiplying the conjugate of the average value by the data after coarse frequency offset correction to obtain the corrected data.

2. The frequency offset estimation and correction method in a severely multipath scattering channel according to claim 1, wherein The expression for the weight of each path is: where C1(k) is the estimated value of the first sub-channel of the k-th path, C2(k) is the estimated value of the second sub-channel of the k-th path, |·| represents the absolute value, and ∑ is the summation symbol.

3. The frequency offset estimation and correction method in a severely multipath scattering channel according to claim 1, characterized in that The local CAZAC sequence includes: a first local CAZAC sequence and a second local CAZAC sequence; The expression of the first sub-channel estimate value for each path is: The expression of the second sub-channel estimate value for each path is: Among them, p1(n) is the first local CAZAC sequence, and p1 * (n) is the conjugate sequence of p1(n), p2(n) is the second local CAZAC sequence, and p2 * (n) is the conjugate sequence of p2(n), f d is the frequency offset to be estimated, ω is the angular frequency corresponding to f d , and ω = 2πf d .

Images