A frequency estimation method for satellite mobile communication system

Abstract

The invention discloses a frequency estimation method for a satellite mobile communication system, and belongs to the field of satellite mobile communication. The method comprises: receiving burst data, performing matched filtering processing; performing square loop timing, performing sample point extraction; performing coarse frequency estimation; performing frequency difference removal operation according to the coarse frequency offset value; performing signal-to-noise ratio estimation, comparing the signal-to-noise ratio with a threshold value P1, if the signal-to-noise ratio is greater than P1, reporting the frequency estimation result, otherwise performing fine frequency estimation; performing frequency difference removal operation again according to the fine frequency offset value; comparing the signal-to-noise ratio with a threshold value P2, if the signal-to-noise ratio is greater than P2, reporting the frequency estimation result, otherwise performing large-step frequency sweeping; if the large-step frequency sweeping is successful, reporting the frequency estimation result, otherwise performing small-step frequency sweeping and reporting the frequency estimation result. The invention can solve the problem that large Doppler frequency deviation and high Doppler frequency deviation change rate lead to a sharp deterioration of receiver performance in a satellite mobile communication system, thereby ensuring reliable service communication.

Description

Technical Field

[0001] The invention belongs to the field of satellite mobile communications, and in particular relates to a frequency estimation method for a satellite mobile communication system. Background Art

[0002] An important feature of satellite communication is its wide coverage. Using multiple satellites can achieve the purpose of global communication. Therefore, the development speed of satellite communication technology has been extremely rapid in the past 20 years. Satellites are usually divided into high-orbit, medium-orbit and low-orbit satellites. High-orbit satellites are in a relatively static state with the ground, and the Doppler frequency shift caused by their own movement is small. However, when the user includes high-speed moving objects such as high-speed trains and airplanes, the Doppler frequency deviation caused by their relative movement is several kHz to tens of kHz. Unlike high-orbit satellites, medium-orbit and low-orbit satellites will cause a large Doppler frequency shift due to their own high-speed movement. If the ground terminal is also a high-speed moving terminal, then during the satellite communication process, when the signal is transmitted through the channel, it will cause a large frequency deviation and phase shift due to the high-speed movement, making it impossible for the signal to be demodulated and decoded normally when it reaches the receiving end. Especially for low-orbit satellites, the moving speed is very fast, and the Doppler frequency shift caused by itself is tens of kHz to several MHz. In high-dynamic scenarios, the Doppler frequency shift is not only large, but also presents high-order time-varying characteristics because the radial velocity and acceleration change continuously over time. Therefore, before normal communication, the received signal must be estimated and compensated to effectively restore the original baseband signal and ensure reliable communication.

[0003] The frequency offset estimation in the prior art is divided into time domain estimation algorithms and frequency domain estimation algorithms. The most basic estimation method based on the time domain is the maximum likelihood estimation algorithm, which has an estimation error that can reach the lower limit of the Cramer-Law bound. It has better performance but has a large amount of calculation and high complexity, and is not suitable for systems with high real-time requirements. However, the accuracy of the frequency offset estimation based on the prior art in the frequency domain is not high. Due to the fence effect in the FFT transform, the Doppler frequency estimation error is relatively large.

[0004] The current typical frequency estimation architecture is a two-stage design of coarse frequency offset estimation plus fine frequency offset estimation. For scenarios with relatively good signal-to-noise ratio, the two-stage frequency estimation scheme has too much computational complexity. For scenarios with relatively poor signal-to-noise ratio, the two-stage frequency estimation scheme cannot accurately estimate the frequency offset value, resulting in probabilistic errors in demodulation and decoding. Summary of the invention

[0005] To solve the problem of frequency estimation accuracy in different scenarios, the present invention proposes a frequency estimation method for a satellite mobile communication system. According to the signal-to-noise ratio estimation result, the present invention determines how many levels of frequency estimation are required, ensuring that under high signal-to-noise ratio conditions, an accurate frequency offset value can be calculated through one-level coarse frequency estimation, and under low signal-to-noise ratio conditions, an accurate frequency offset value can be calculated through multi-level frequency estimation. This not only ensures a sufficiently large frequency estimation range but also ensures that the frequency estimation value is sufficiently accurate, and can adapt to the characteristics of large Doppler frequency shift and high Doppler frequency change rate in the high-dynamic environment of the satellite mobile communication system, ensuring that the received signal can be correctly demodulated and decoded in any scenario.

[0006] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0007] A frequency estimation method for a satellite mobile communication system, comprising the following steps:

[0008] Step 1, the receiving end receives and stores burst data, and performs matched filtering processing on the stored burst data;

[0009] Step 2, perform square loop timing on the data after matched filtering in Step 1, and according to the timing result, perform sample extraction processing on the data after matched filtering to obtain symbol data;

[0010] Step 3, perform coarse frequency estimation on the symbol data to obtain a coarse frequency offset value freq_est_coarse;

[0011] Step 4, perform a frequency difference removal operation on the symbol data according to the coarse frequency offset value;

[0012] Step 5, perform signal-to-noise ratio estimation on the data after frequency difference removal in Step 4, compare the estimated signal-to-noise ratio with a threshold P1. If the signal-to-noise ratio is greater than P1, report the frequency estimation result freq_est = freq_est_coarse, and complete the frequency estimation. Otherwise, continue with the subsequent steps;

[0013] Step 6, perform fine frequency estimation on the data after frequency difference removal in Step 4 to obtain a fine frequency offset value freq_est_fine;

[0014] Step 7, perform a frequency difference removal operation on the data after frequency difference removal in Step 4 again according to the fine frequency offset value;

[0015] Step 8, compare the signal-to-noise ratio estimated in Step 5 with a threshold P2, P2 < P1. If the signal-to-noise ratio is greater than P2, report the frequency estimation result freq_est = freq_est_coarse + freq_est_fine, and complete the frequency estimation. Otherwise, continue with the subsequent steps;

[0016] Step 9, perform large-step frequency sweeping within the frequency sweeping range, wherein, at each large-step frequency sweeping frequency, process the data after the frequency difference is removed again in step 7 to obtain decoded data. If the CRC check value of the decoded data is correct, report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+current large-step frequency sweeping frequency to complete the frequency estimation. Otherwise, continue to perform large-step frequency sweeping. If the CRC check value of the decoded data is still wrong until the end of the large-step frequency sweeping, continue to the subsequent steps.

[0017] Step 10, perform small-step frequency sweeping within the sweeping range, wherein, at each small-step frequency sweeping frequency, process the data after frequency difference removal in step 7 to obtain decoded data. If the CRC check value of the decoded data is correct or the small-step frequency sweeping is completed, the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+current small-step frequency sweeping frequency is reported to complete the frequency estimation, otherwise the small-step frequency sweeping is continued.

[0018] Furthermore, the specific method of square ring timing in step 2 is:

[0019]

[0020] Where N represents the oversampling multiple, arg(.) represents the angle value, and L represents the total length of the data after matched filtering. Represents the sample data numbered n in the matched filtered data, Indicates the timing result, that is, the timing error sample value.

[0021] Furthermore, in step 3, the coarse frequency offset value is calculated as follows:

[0022]

[0023] in, Indicates the data obtained by modulating the symbol data to the Mth power and then filling the modulated data with 0. For BPSK modulation, M=2, for QPSK modulation, M=4, and for 8PSK modulation, M=8; is the number of FFT points obtained by raising the symbol data length to the power of 2. FFT(.) represents the fast Fourier transform. Indicates the position corresponding to the maximum value of the data. is the symbol rate.

[0024] Furthermore, the threshold value P1 is greater than 10, and the threshold value P2 is in the range of 3 to 8.

[0025] Furthermore, in step 6, the fine frequency offset value is calculated as follows:

[0026]

[0027] in, is the symbol rate, arg(.) indicates the angle value, conj(.) indicates the conjugate value, r(n) is the data after frequency difference removal in step 4, uw1(n) is the data of the first segment unique code, uw2(n) is the data of the second segment unique code, The position information of the middle symbol of the first segment unique code. is the position information of the middle symbol of the second segment unique code, n1 and n2 are the start position subscript and end position subscript of the first segment unique code in the burst data, and n3 and n4 are the start position subscript and end position subscript of the second segment unique code in the burst data.

[0028] Furthermore, the specific method of step 9 is:

[0029] Step 9-1, set the frequency sweep range to (-freq,freq), the frequency sweep step to f_big_step, and initialize the number of iterations n=1;

[0030] Step 9-2, calculate the current large step sweep frequency freq_set_big=-freq+(n-1)*f_big_step;

[0031] Step 9-3, generating a single tone signal of the current large-step sweep frequency, multiplying the data after the frequency difference is removed again in step 7 by the single tone signal to obtain data after further frequency difference removal;

[0032] Step 9-4, performing phase estimation on the data after further frequency difference removal, and completing the initial phase removal operation;

[0033] Step 9-5, soft demodulating the data after the initial phase is removed, outputting soft information, and using the soft information to complete the decoding operation;

[0034] Step 9-6, calculate the CRC check value of the decoded data. If the CRC check value is correct, report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+freq_set_big to complete the frequency estimation. Otherwise, set n=n+1 and return to step 9-2. If the CRC check value of the decoded data is still wrong until (n-1)*f_big_step=freq, continue to execute step 10.

[0035] Furthermore, the specific method of step 10 is:

[0036] Step 10-1, set the frequency sweep range to (-freq,freq), the frequency sweep step to f_small_step, and initialize the number of iterations n=1;

[0037] Step 10-2, calculate the current small step sweep frequency freq_set_small=-freq+(n-1)*f_small_step;

[0038] Step 10-3, generating a single tone signal of the current small step sweep frequency, multiplying the data after the frequency difference is removed again in step 7 by the single tone signal to obtain data after further frequency difference removal;

[0039] Step 10-4, performing phase estimation on the data after further frequency difference removal, and completing the initial phase removal operation;

[0040] Step 10-5, soft demodulating the data after the initial phase is removed, outputting soft information, and using the soft information to complete the decoding operation;

[0041] Step 10-6, calculate the CRC check value of the decoded data. If the CRC check value is correct, report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+freq_set_small to complete the frequency estimation. Otherwise, set n=n+1 and return to step 10-2. If the CRC check value of the decoded data is still wrong until (n-1)*f_big_step=freq, report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+freq to complete the frequency estimation.

[0042] The beneficial effects achieved by adopting the above technical solution are:

[0043] 1. The present invention can solve the problem that the performance of the receiver is sharply deteriorated due to large Doppler frequency shift and high Doppler change rate in a high dynamic environment. The present invention can accurately estimate the Doppler frequency deviation value of the received signal, and can demodulate and decode the received signal regardless of the signal-to-noise ratio in which range, thereby ensuring reliable communication of the satellite mobile communication system.

[0044] 2. The present invention is implemented in a pure software manner and does not rely on special hardware. When the received signal-to-noise ratio is very large, an accurate Doppler frequency offset value can be obtained through coarse frequency estimation, thereby reducing the amount of calculation. When the received signal-to-noise ratio is very low, an accurate frequency offset value can be obtained through fine frequency estimation or large-step frequency sweeping or small-step frequency sweeping, thereby ensuring that an accurate frequency offset value can still be output under a low signal-to-noise ratio.

[0045] 3. The present invention can select a single-stage frequency estimation method or a multi-stage frequency estimation method according to different signal-to-noise ratio conditions of the application and the limitation of the equipment processing capacity, so as to facilitate the use in different application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 The present invention is a block diagram of the transmitter component of a frequency estimation method for a satellite mobile communication system.

[0047] Figure 2 It is a schematic diagram of the business and control burst signal structure of a mobile communication system.

[0048] Figure 3 It is a flow chart of frequency estimation at the receiving end in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The technical solution of the present invention is described in detail below in conjunction with the accompanying drawings.

[0050] A frequency estimation method for a satellite mobile communication system comprises the following steps:

[0051] Step 1, the receiving end receives and stores burst data, and performs matched filtering on the stored burst data;

[0052] Step 2, performing square loop timing on the data after matched filtering in step 1, and performing sample point extraction processing on the data after matched filtering according to the timing result to obtain symbol data; the specific method of square loop timing is:

[0053]

[0054] Where N represents the oversampling multiple, arg(.) represents the angle value, and L represents the total length of the data after matched filtering. Represents the sample data numbered n in the matched filtered data, Indicates the timing result, that is, the timing error sample value.

[0055] Step 3, perform coarse frequency estimation on the symbol data to obtain a coarse frequency offset value freq_est_coarse; the coarse frequency offset value is calculated as follows:

[0056]

[0057] in, Indicates the data obtained by modulating the symbol data to the Mth power and then filling the modulated data with 0. For BPSK modulation, M=2, for QPSK modulation, M=4, and for 8PSK modulation, M=8; is the number of FFT points obtained by raising the symbol data length to the power of 2. FFT(.) represents the fast Fourier transform. Indicates the position corresponding to the maximum value of the data. is the symbol rate.

[0058] Step 4, performing a frequency difference removal operation on the symbol data according to the coarse frequency offset value;

[0059] Step 5, estimate the signal-to-noise ratio of the data after the frequency difference is removed in step 4, and compare the estimated signal-to-noise ratio with the threshold P1. The threshold P1 is greater than 10. If the signal-to-noise ratio is greater than P1, report the frequency estimation result freq_est=freq_est_coarse, and complete the frequency estimation. Otherwise, continue with the subsequent steps.

[0060] Step 6: Perform fine frequency estimation on the data after the frequency difference is removed in step 4 to obtain a fine frequency offset value freq_est_fine. The fine frequency offset value is calculated as follows:

[0061]

[0062] in, is the symbol rate, arg(.) indicates the angle value, conj(.) indicates the conjugate value, r(n) is the data after frequency difference removal in step 4, uw1(n) is the data of the first segment unique code, uw2(n) is the data of the second segment unique code, The position information of the middle symbol of the first segment unique code. is the position information of the middle symbol of the second segment unique code, n1 and n2 are the start position subscript and end position subscript of the first segment unique code in the burst data, and n3 and n4 are the start position subscript and end position subscript of the second segment unique code in the burst data.

[0063] Step 7, performing frequency difference removal operation again on the data after frequency difference removal in step 4 according to the fine frequency offset value;

[0064] Step 8: compare the signal-to-noise ratio estimated in step 5 with the threshold value P2, where the threshold value P2 ranges from 3 to 8. If the signal-to-noise ratio is greater than P2, report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine to complete the frequency estimation, otherwise continue with the subsequent steps.

[0065] Step 9, perform large-step frequency sweeping within the frequency sweeping range, wherein, at each large-step frequency sweeping frequency, process the data after the frequency difference is removed again in step 7 to obtain decoded data. If the CRC check value of the decoded data is correct, report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+current large-step frequency sweeping frequency to complete the frequency estimation, otherwise continue to perform large-step frequency sweeping. If the CRC check value of the decoded data is still wrong until the end of the large-step frequency sweeping, continue to the subsequent steps; the specific method is:

[0066] Step 9-1, set the frequency sweep range to (-freq,freq), the frequency sweep step to f_big_step, and initialize the number of iterations n=1;

[0067] Step 9-2, calculate the current large step sweep frequency freq_set_big=-freq+(n-1)*f_big_step;

[0068] Step 9-3, generating a single tone signal of the current large-step sweep frequency, multiplying the data after the frequency difference is removed again in step 7 by the single tone signal to obtain data after further frequency difference removal;

[0069] Step 9-4, performing phase estimation on the data after further frequency difference removal, and completing the initial phase removal operation;

[0070] Step 9-5, soft demodulating the data after the initial phase is removed, outputting soft information, and using the soft information to complete the decoding operation;

[0071] Step 9-6, calculate the CRC check value of the decoded data. If the CRC check value is correct, report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+freq_set_big to complete the frequency estimation. Otherwise, set n=n+1 and return to step 9-2. If the CRC check value of the decoded data is still wrong until (n-1)*f_big_step=freq, continue to execute step 10.

[0072] Step 10, perform small step frequency sweep within the sweep range, wherein, at each small step frequency sweep frequency, process the data after frequency difference removal in step 7 again to obtain decoded data, if the CRC check value of the decoded data is correct or the small step frequency sweep is completed, then report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+current small step frequency sweep frequency, and complete the frequency estimation, otherwise continue to perform small step frequency sweep. The specific method is:

[0073] Step 10-1, set the frequency sweep range to (-freq,freq), the frequency sweep step to f_small_step, and initialize the number of iterations n=1;

[0074] Step 10-2, calculate the current small step sweep frequency freq_set_small=-freq+(n-1)*f_small_step;

[0075] Step 10-3, generating a single tone signal of the current small step sweep frequency, multiplying the data after the frequency difference is removed again in step 7 by the single tone signal to obtain data after further frequency difference removal;

[0076] Step 10-4, performing phase estimation on the data after further frequency difference removal, and completing the initial phase removal operation;

[0077] Step 10-5, soft demodulating the data after the initial phase is removed, outputting soft information, and using the soft information to complete the decoding operation;

[0078] Step 10-6, calculate the CRC check value of the decoded data. If the CRC check value is correct, report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+freq_set_small to complete the frequency estimation. Otherwise, set n=n+1 and return to step 10-2. If the CRC check value of the decoded data is still wrong until (n-1)*f_big_step=freq, report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+freq to complete the frequency estimation.

[0079] Here is a more specific example:

[0080] A frequency estimation method for a satellite mobile communication system, the sending end is composed of a block diagram as shown in Figure 1 As shown, the original information is transmitted after data packaging, CRC addition, encoding, unique code addition, protection symbol addition, modulation, shaping filtering, and up-conversion.

[0081] The structure of business and control burst signals in mobile communication systems is as follows: Figure 2 As shown, it specifically includes four parts: a protection symbol at the beginning, a unique code, information, and a protection symbol at the end of the burst.

[0082] Based on this transmission block diagram and burst structure, in order to realize frequency estimation, it is necessary to perform coarse frequency estimation, fine frequency estimation, sweep frequency estimation and other processing processes, such as Figure 3 As shown, the specific steps include:

[0083] S101, a receiving end stores received burst sample data according to a receiving time;

[0084] S102, performing matched filtering on the received burst sample data;

[0085] S103, the data after matched filtering is processed , perform square loop timing, and perform sample point extraction processing based on the timing result to convert the sample point data into symbol data; calculate the timing value according to the following formula.

[0086] (1)

[0087] Where N represents the oversampling multiple, arg(.) represents the angle value of the data, and L represents the total length of the received samples. Indicates the timing error sample value. The best sample point is extracted from the received sample point according to the timing error value, and then the best symbol is obtained.

[0088] S104, performing a coarse frequency estimation on the extracted symbol data, and performing M-th power demodulation on the symbol data, wherein M=2 for BPSK modulation, M=4 for QPSK modulation, and M=8 for 8PSK modulation;

[0089] Fill the de-modulated data with 0, perform FFT calculation, calculate the absolute value of the result of FFT calculation, and find the position of the data with the maximum absolute value, and convert the position into coarse frequency information freq_est_coarse, and store the frequency;

[0090] (2)

[0091] in, is the burst symbol rate, It means that when the maximum value is obtained for the data, the position information corresponding to the maximum value is obtained. FFT(.) means that the FFT calculation is performed on the data. The number of FFT points obtained by raising the received symbol data length to the power of 2.

[0092] S105: Remove the frequency difference in the received burst symbol according to the coarse frequency estimation information.

[0093] S106, estimating the signal-to-noise ratio of the symbol data from which the coarse frequency is removed, to obtain the signal-to-noise ratio of the received burst data.

[0094] S107, determine the signal-to-noise ratio data SNR, if SNR>P1, where P1 can be obtained based on simulation results according to different channel structures, modulation methods, coding methods, etc., in this embodiment, P1=15dB, then go to step S108, if SNR≤P1, go to step S109;

[0095] S108, calculate the estimated frequency, the frequency is freq_est=freq_est_coarse, and go to step S129;

[0096] S109, performing fine frequency estimation again on the data after removing the coarse frequency, extracting the unique code data in the burst data, performing correlation processing on the received pilot data and the transmitted data, and calculating the fine frequency estimation information according to the phase information of the unique code data, as shown in the following formula;

[0097] (3)

[0098] in, is the burst symbol rate, arg(.) indicates the angle value of the data, conj(.) indicates the conjugate value of the data, r(n) is the data to be processed, uw1(n) is the data of the first segment of the unique code, uw2(n) is the data of the second segment of the unique code, The position information of the middle symbol of the first segment unique code. The position information of the middle symbol of the second segment unique code;

[0099] S110, removing fine frequencies in received burst symbols according to the fine frequency estimation information;

[0100] S111, determine the signal-to-noise ratio data SNR, if SNR>P2, where P2 can be obtained based on simulation results according to different channel structures, modulation methods, coding methods, etc., in this embodiment, P2=5dB, then go to step S112, if SNR≤P2, go to step S113;

[0101] S112, calculate the estimated frequency, the frequency is freq_est=freq_est_coarse+freq_est_fine, and go to step S129;

[0102] S113, perform large-step frequency sweep calculation, set the frequency sweep step to f_big_step, the frequency sweep range to (-freq, freq), initialize the large frequency sweep number n=1, in this embodiment, set freq= 50Hz, f_big_step=10Hz, start from -freq, perform signal processing each time with a step of f_big_step, it can be seen that the frequency of each frequency sweep freq_set_big= -freq+(n-1)*f_big_step, and a total of 11 times are required to complete the entire frequency sweep;

[0103] S114, the frequency sweep number n=n+1, the frequency sweep frequency is calculated according to the current frequency sweep number, and freq_set_big in the received burst symbol is removed, as shown in the following formula;

[0104] (4)

[0105] in, To remove the data of the current frequency difference, r(n) is the input data, is the receive symbol period.

[0106] S115, obtaining received data with the frequency sweep information removed, estimating the phase information of the received burst, and removing the initial phase information;

[0107] (5)

[0108] in, is the estimated phase information, exp(.) is the exponential processing of the data, arg(.) means the angle value of the data, r(n) is the data to be processed, conj(.) means the conjugate value of the data, and uw1(n) is the data of the first segment of the unique code.

[0109] The initial phase can be removed by conjugating and multiplying the received data and the phase information;

[0110] (6)

[0111] in, To remove the phase data.

[0112] S116, extracting coding information from the received data according to the burst structure, and performing soft demodulation;

[0113] S117, decoding the soft demodulated data;

[0114] S118, decode the CRC of the decoded data, if the CRC check is correct, go to step S119; if the CRC check is wrong, go to step S120;

[0115] S119, calculate the estimated frequency, the frequency is freq_est=freq_est_coarse+freq_est_fine+freq_set_big, go to step S129;

[0116] S120, determining whether the large-step frequency sweep is completed, if the frequency sweep is completed, proceeding to step S121, if the frequency sweep is not completed, proceeding to step S114;

[0117] S121, perform small-step frequency sweep calculation, set the frequency sweep step to f_small_step, frequency sweep range (-freq,freq), initialize the small frequency sweep times n=1, where freq=50Hz, f_small_step=1Hz, start from -freq, perform signal processing with each step f_small_step, it can be seen that the frequency of each frequency sweep freq_set_small= -freq+ (n-1)* f_small_step, and a total of 101 times are required to complete the entire frequency sweep;

[0118] S122, the number of sweep times n=n+1, the sweep frequency is calculated according to the current number of sweep times, and freq_set_small in the received burst symbol is removed, as shown in the following formula;

[0119] (7)

[0120] in, To remove the data of the current frequency difference, r(n) is the input data, is the receive symbol period.

[0121] S123, obtaining the unique code data of the received data after removing the frequency sweep information, estimating the phase information of the received burst, and completing the initial phase removal; the specific scheme is shown in step S115;

[0122] S124, extracting coding information from the received data according to the burst structure, and performing soft demodulation;

[0123] S125, decoding the soft demodulated data;

[0124] S126, decode the CRC of the decoded data, if the CRC check is correct, go to step S127; if the CRC check is wrong, go to step S128;

[0125] S127, calculate the estimated frequency, the frequency is freq_est=freq_est_coarse+freq_est_fine+freq_set_small, and go to step S129;

[0126] S128, determining whether the small frequency sweep times are completed, if the frequency sweep is completed, proceeding to step S112, if the frequency sweep is not completed, proceeding to step S122;

[0127] S129, reporting the estimated frequency freq_est, and completing the frequency estimation.

[0128] The present invention can solve the problem that a large Doppler frequency deviation and a high Doppler frequency deviation change rate lead to a sharp deterioration of receiver performance in a satellite mobile communication system, thereby ensuring reliable business communications.

Claims

1. A frequency estimation method for a satellite mobile communication system, characterized in that: It includes the following steps: Step 1: The receiving end receives and stores the burst data, and performs matched filtering processing on the stored burst data; Step 2: Perform square loop timing on the data after matched filtering in Step 1. According to the timing result, perform sample extraction processing on the data after matched filtering to obtain symbol data; Step 3: Perform coarse frequency estimation on the symbol data to obtain the coarse frequency offset value freq_est_coarse; Step 4: Perform frequency offset removal operation on the symbol data according to the coarse frequency offset value; Step 5: Perform signal-to-noise ratio estimation on the data after frequency offset removal in Step 4, compare the estimated signal-to-noise ratio with the threshold P1. If the signal-to-noise ratio is greater than P1, report the frequency estimation result freq_est = freq_est_coarse, and complete the frequency estimation. Otherwise, continue with the subsequent steps; Step 6: Perform fine frequency estimation on the data after frequency offset removal in Step 4 to obtain the fine frequency offset value freq_est_fine; Step 7: According to the fine frequency offset value, perform frequency offset removal operation on the data after frequency offset removal in Step 4 again; Step 8: Compare the signal-to-noise ratio estimated in Step 5 with the threshold P2, P2 < P1. If the signal-to-noise ratio is greater than P2, report the frequency estimation result freq_est = freq_est_coarse + freq_est_fine, and complete the frequency estimation. Otherwise, continue with the subsequent steps; Step 9: Perform large-step frequency sweeping within the frequency sweeping range. Among them, at each large-step frequency sweeping frequency, process the data after frequency offset removal again in Step 7 to obtain decoded data. If the CRC check value of the decoded data is correct, report the frequency estimation result freq_est = freq_est_coarse + freq_est_fine + the current large-step frequency sweeping frequency, and complete the frequency estimation. Otherwise, continue with the large-step frequency sweeping. If the CRC check value of the decoded data is still incorrect until the end of the large-step frequency sweeping, continue with the subsequent steps; Step 10: Perform small-step frequency sweeping within the frequency sweeping range. Among them, at each small-step frequency sweeping frequency, process the data after frequency offset removal again in Step 7 to obtain decoded data. If the CRC check value of the decoded data is correct or the small-step frequency sweeping ends, report the frequency estimation result freq_est = freq_est_coarse + freq_est_fine + the current small-step frequency sweeping frequency, and complete the frequency estimation. Otherwise, continue with the small-step frequency sweeping.

2. The frequency estimation method of a satellite mobile communication system according to claim 1, characterized in that: The specific method of square loop timing in Step 2 is: Where N represents the oversampling multiple, arg(.) represents the angle value, and L represents the total length of the data after matched filtering. Represents the sample data numbered n in the matched filtered data, Indicates the timing result, that is, the timing error sample value.

3. The frequency estimation method of a satellite mobile communication system according to claim 2, characterized in that: In Step 3, the calculation method of the coarse frequency offset value is: in, Indicates the data obtained by modulating the symbol data to the Mth power and then filling the modulated data with 0. For BPSK modulation, M=2, for QPSK modulation, M=4, and for 8PSK modulation, M=8; is the number of FFT points obtained by raising the symbol data length to the power of 2. FFT(.) represents the fast Fourier transform. Indicates the position corresponding to the maximum value of the data. is the symbol rate.

4. The frequency estimation method of a satellite mobile communication system according to claim 3, characterized in that: The threshold P1 is greater than 10, and the value range of the threshold P2 is between 3 and 8.

5. The frequency estimation method of a satellite mobile communication system according to claim 4, characterized in that: In Step 6, the calculation method of the fine frequency offset value is: in, is the symbol rate, arg(.) indicates the angle value, conj(.) indicates the conjugate value, r(n) is the data after frequency difference removal in step 4, uw1(n) is the data of the first segment unique code, uw2(n) is the data of the second segment unique code, The position information of the middle symbol of the first segment unique code. is the position information of the middle symbol of the second segment unique code, n1 and n2 are the start position subscript and end position subscript of the first segment unique code in the burst data, and n3 and n4 are the start position subscript and end position subscript of the second segment unique code in the burst data.

6. The frequency estimation method of a satellite mobile communication system according to claim 5, characterized in that: The specific method of Step 9 is: Step 9-1: Set the frequency sweeping range as (-freq, freq), the frequency sweeping step as f_big_step, and initialize the iteration number n = 1; Step 9-2: Calculate the current large-step frequency sweeping frequency freq_set_big = -freq + (n - 1) * f_big_step; Step 9-3, generating a single tone signal of the current large-step sweep frequency, multiplying the data after the frequency difference is removed again in step 7 by the single tone signal to obtain data after further frequency difference removal; Step 9-4, performing phase estimation on the data after further frequency difference removal, and completing the initial phase removal operation; Step 9-5, soft demodulating the data after the initial phase is removed, outputting soft information, and using the soft information to complete the decoding operation; Step 9-6, calculate the CRC check value of the decoded data. If the CRC check value is correct, report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+freq_set_big to complete the frequency estimation. Otherwise, set n=n+1 and return to step 9-2. If the CRC check value of the decoded data is still wrong until (n-1)*f_big_step=freq, continue to execute step 10.

7. The frequency estimation method of a satellite mobile communication system according to claim 6, characterized in that: The specific method of step 10 is: Step 10-1, set the frequency sweep range to (-freq,freq), the frequency sweep step to f_small_step, and initialize the number of iterations n=1; Step 10-2, calculate the current small step sweep frequency freq_set_small=-freq+(n-1)*f_small_step; Step 10-3, generating a single tone signal of the current small step sweep frequency, multiplying the data after the frequency difference is removed again in step 7 by the single tone signal to obtain data after further frequency difference removal; Step 10-4, performing phase estimation on the data after further frequency difference removal, and completing the initial phase removal operation; Step 10-5, soft demodulating the data after the initial phase is removed, outputting soft information, and using the soft information to complete the decoding operation; Step 10-6, calculate the CRC check value of the decoded data. If the CRC check value is correct, report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+freq_set_small to complete the frequency estimation. Otherwise, set n=n+1 and return to step 10-2. If the CRC check value of the decoded data is still wrong until (n-1)*f_big_step=freq, report the frequency estimation result freq_est=freq_est_coarse+freq_est_fine+freq to complete the frequency estimation.

Images