Rapid differential frame synchronization method under burst transmission system

Abstract

The invention provides a fast differential frame synchronization method under a burst transmission system, which comprises the following steps: a receiver reduces the sampling rate of a burst transmission sequence to four times of a symbol rate, detects the timing error of a down-sampling sequence, calculates an integer parameter and a fraction parameter based on a timing error result, and performs frame synchronization according to the integer parameter and the fraction parameter; then, a cubic interpolation filter is used for calibrating timing errors and recovering a correct data sequence, a differential frame synchronization method which is insensitive to signal amplitude and suitable for any phase deviation is used for searching a frame header position, and whether the size of a correlation peak value is larger than a set threshold value or not is judged; if yes, timing synchronization and frame synchronization are completed; and if not, returning to continue to calculate the timing error. The method is suitable for a burst transmission system, is different from a traditional frame synchronization method, does not need to carry out frequency difference estimation, amplitude control and carrier synchronization in advance, and can quickly realize timing synchronization and frame synchronization, so that the synchronization time is shortened.

Description

Technical Field

[0001] The invention relates to the field of communication technology, and in particular to a fast differential frame synchronization method under a burst transmission system. Background Art

[0002] Burst transmission, also generally referred to as data burst, generally refers to relatively high-bandwidth data transmission in a short period of time in the communication field.

[0003] Chinese patent publication number CN118249973A, publication date is June 25, 2024, named "Method, device and computing equipment for timing synchronization in digital communication system", which discloses a method of increasing the accuracy of synchronization position by using a secondary verification method. The method is characterized in that the received signal is segmented and differential accumulation is performed, the synchronization position is determined based on the result and the local differential sequence, and the synchronized sequence to be verified is Fourier transformed to determine the frequency deviation. This is a timing synchronization method that increases the capture probability by increasing the number of verifications, but its shortcomings are: this method has high requirements on the signal transmission system, and cannot meet the needs of fast capture and tracking of short pilot burst signal sequences. The segmented differential accumulation calculation of the received signal requires a sufficiently long fixed pilot symbol, and the burst transmission system signal sequence length is not enough to support secondary verification operations.

[0004] Chinese patent publication number CN117615047A, published on February 27, 2024, entitled "A design method for a Beidou short message receiver based on FPGA" discloses a method for achieving frame header capture through timing control and tracking through a loop. The method is characterized in that the short message sampling data buffer module is combined with the capture module, and two-way capture parallel timing control is designed to achieve the synchronization process of the frame header. This method is a synchronization method with high capture accuracy and continuous tracking, but its disadvantage is that the stabilization of the loop tracking module in this method takes time and is not suitable for the capture and tracking of short-term burst signals.

[0005] Therefore, a burst transmission method that can quickly capture tracking requirements is needed. Summary of the invention

[0006] The present invention is to solve the problems of long synchronization time, high probability of missed capture and high probability of false capture under the burst transmission system, and provides a fast differential frame synchronization method under the burst transmission system. The receiver reduces the sampling rate of the burst transmission sequence to 4 times the symbol rate, detects the timing error of the down-sampling sequence, calculates integer parameters and fractional parameters based on the timing error result, and then uses a cubic interpolation filter to calibrate the timing error and restore the correct data sequence. The frame header position is found by using a differential frame synchronization method that is insensitive to the signal amplitude and applicable to any phase deviation, and determines whether the relevant peak value is greater than the set threshold; if it is satisfied, the timing synchronization and frame synchronization are completed; if it is not satisfied, it returns to continue calculating the timing error. The present invention is applicable to the burst transmission system, and is different from the traditional frame synchronization method. It does not need to perform frequency difference estimation, amplitude control, and carrier synchronization in advance, and can quickly achieve timing synchronization and frame synchronization, thereby shortening the synchronization time.

[0007] The present invention provides a fast differential frame synchronization method under a burst transmission system, comprising the following steps:

[0008] S1. The receiver performs matched filtering and downsampling on the received burst transmission sequence, and reduces the sampling rate to N times the symbol rate to obtain a downsampled sequence. The downsampled sequence includes a pilot sequence and a data sequence. The total length of the downsampled sequence is K, and the pilot sequence is of length K. p , a 0 / 1 sequence with fixed content; the data sequence is of length K d , 0 / 1 sequence with random content;

[0009] The down-sampling sequence x(n) is obtained by sliding sampling at fixed intervals, where n is initialized to 1. Each down-sampling sequence includes NL0 sampling points, and the sampling time difference of the first sampling points of two adjacent down-sampling sequences is N. L0 is the observation interval length;

[0010] S2, get the timing error of the down-sampling sequence x(n)

[0011]

[0012] Where k is the sampling point position index in the data sequence of the down-sampled sequence x(n), T s is the sampling period of the phase-finding operator, arg(·) is the phase-finding operator, and T is the period of the phase-finding operator;

[0013] S3, based on timing error Get the integer parameter m(n) and fractional parameter μ(n) of the downsampled sequence x(n):

[0014]

[0015] S4, using a cubic interpolation filter to calibrate the timing error and recover the correct data sequence y(u);

[0016] Cubic interpolation filter coefficients c i for:

[0017] c -2 (μ(n))=μ(n) 3 / 6-μ(n) / 6

[0018] c -1 (μ(n))=-μ(n) 3 / 2+μ(n) 2 / 2+μ(n)

[0019] c0(μ(n))=μ(n) 3 / 2-μ(n) 2 -μ(n) / 2+1

[0020] c1(μ(n))=-μ(n) 3 / 6+μ(n) 2 / 2-μ(n) / 3;

[0021] The cubic interpolation filtering process is:

[0022]

[0023] Determine whether the starting position N of the data frame header is obtained F If yes, n=n+1, return to step S2, if no, go to step S5;

[0024] S5, selecting a data sequence y(u) of a specified length and performing a differential frame synchronization operation on the preceding and following symbols in the sequence to obtain a differential frame synchronization result R(u);

[0025]

[0026] Among them, p h is part of the pilot symbol in the physical layer frame, with a length of K sp , K sp ≤K p ;p u is a pilot sequence of the same length as the data sequence y(u) of the specified length; sign(·) is a sign function used to obtain the sign bit of the symbol in the data sequence; real(·) is the real part of the complex signal; imag(·) is the imaginary part of the complex signal; (·) * It is a conjugate calculation operation;

[0027] S6. Compare the differential frame synchronization result R(u) with the threshold R T For comparison, when |R(u)|>R TWhen u is the starting position of the data frame header, marked as N F ;

[0028] n=n+1, return to step S2;

[0029] Until all burst transmission sequence processing is completed, a fast differential frame synchronization method under a burst transmission system is completed.

[0030] The fast differential frame synchronization method under a burst transmission system described in the present invention is, as a preferred embodiment, in step S1, N is 4, the sampling rate is 4 times; the observation interval length is related to the timing error estimation accuracy;

[0031] The fast differential frame synchronization method under the burst transmission system described in the present invention is, as a preferred method, when the number of sampling points in the down-sampling sequence x(n) in step S1 is less than NL0, the timing error is used The integer parameter m(n-1) and the fractional parameter μ(n-1) perform cubic interpolation filtering in step S4.

[0032] The fast differential frame synchronization method under a burst transmission system described in the present invention is, as a preferred mode, to re-perform the sampling in step S1 if a communication interruption occurs during the communication process.

[0033] The fast differential frame synchronization method under the burst transmission system described in the present invention is, as a preferred method, L0≤KK p .

[0034] The fast differential frame synchronization method under a burst transmission system described in the present invention is, as a preferred method, in step S1, the receiver simultaneously receives synchronization information, the synchronization information includes: sending time slot number, receiving time slot number, TOD, frequency hopping time difference and frequency hopping rules; the frequency hopping rules include: frequency hopping speed, frequency hopping block size and number of frequency hopping blocks per unit time slot.

[0035] The fast differential frame synchronization method under the burst transmission system described in the present invention is, as a preferred embodiment, in step S6, the threshold R T It is set according to the demodulation threshold, missed alarm probability and false alarm probability. If normalization is performed, the threshold R T Not greater than the length of the pilot part selected in step S5; if normalization is not performed, it is necessary to simulate and set an appropriate threshold value according to the demodulation threshold requirements of the actual project.

[0036] The fast differential frame synchronization method under a burst transmission system described in the present invention, as a preferred mode, uses a timing synchronization device based on a time slot frequency hopping network, and the timing synchronization device includes a transmitter processor and a receiver processor.

[0037] In the fast differential frame synchronization method under a burst transmission system described in the present invention, as a preferred embodiment, both the sending processor and the receiving processor are FPGAs.

[0038] The present invention has the following advantages:

[0039] (1) The present invention is applicable to the burst transmission system and is different from the traditional frame synchronization method. It does not require frequency difference estimation, amplitude control, and carrier synchronization in advance, and can quickly achieve timing synchronization and frame synchronization, thereby shortening the synchronization time.

[0040] (2) Since the correlation calculation of differential frame synchronization is to perform differential correlation on the symbol after taking the sign bit, the influence of the symbol amplitude on the synchronization result is avoided, and the probability of false capture is reduced;

[0041] (3) Since the correlation calculation of differential frame synchronization is to first perform soft information differential correlation and then take the sign bit of the differential correlation result, the influence of special phase deviation on the synchronization result is avoided and the probability of missed capture is reduced.

[0042] (4) The software independently implements frame synchronization, and no special equipment is required to maintain the initial time and timing system of the transmitter and receiver, which reduces the cost of the product.

[0043] (5) The hardware platform is highly versatile and the modules are highly portable.

[0044] (6) Applicable to various communication environments, including scenarios with long-distance, short-distance, and high-mobility characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A flow chart of a fast differential frame synchronization method under a burst transmission system;

[0046] Figure 2 A flow chart of a fast differential frame synchronization method under a burst transmission system;

[0047] Figure 3 It is a schematic diagram of a peak detection curve of a correlation result in Example 2 of a fast differential frame synchronization method under a burst transmission system;

[0048] Figure 4 A schematic diagram of the peak value and frame header position of the differential frame synchronization result with different demodulation thresholds in Example 2 of a fast differential frame synchronization method under a burst transmission system;

[0049] Figure 5 is the effect of different demodulation thresholds on the differential frame synchronization result in Example 2 of a fast differential frame synchronization method under a burst transmission system. b / N0 and bit error rate influence characteristic diagram;

[0050] Figure 6 A schematic diagram of frequency deviation in Embodiment 2 of a fast differential frame synchronization method under a burst transmission system;

[0051] Figure 7 It is a simulation schematic diagram of the differential frame synchronization results of different frequency offsets in Example 2 of a fast differential frame synchronization method under a burst transmission system. DETAILED DESCRIPTION

[0052] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0053] Example 1

[0054] like Figures 1-2 As shown, a fast differential frame synchronization method under a burst transmission system includes the following steps:

[0055] S1. The receiver performs matched filtering and downsampling on the received burst transmission sequence, reducing the sampling rate of the burst sequence to 4 times the symbol rate. The burst sequence contains a pilot part and a data part, where the pilot part is K long. p The pilot content is a fixed 0 / 1 sequence with good autocorrelation characteristics; the length of the data part is K d , the data content is a random 0 / 1 sequence;

[0056] S2. Calculate the timing error of the burst sequence x(n) in step S1:

[0057]

[0058] Where arg(·) represents the phase finding operator; k is the position index of the sampling point in the data sequence; T s represents the sampling period of the symbol; T represents the symbol period; N represents the number of sampling points in a data symbol period. This method requires a value of 4, corresponding to the 4 times sampling rate in step (1); L0 represents the length of the observation interval. The value of this length will affect the accuracy of the timing error estimation.

[0059] S3, based on the calculation results of the timing error Compute the integer parameter m(n) and the fractional parameter μ(n):

[0060]

[0061] S4. Use cubic interpolation filter to calibrate timing error and recover the correct data sequence:

[0062] The cubic (i=-2,-1,0,1) interpolation filter coefficients are:

[0063] c -2 (μ(n))=μ(n) 3 / 6-μ(n) / 6

[0064] c -1 (μ(n))=-μ(n) 3 / 2+μ(n) 2 / 2+μ(n)

[0065] c0(μ(n))=μ(n) 3 / 2-μ(n) 2 -μ(n) / 2+1

[0066] c1(μ(n))=-μ(n) 3 / 6+μ(n) 2 / 2-μ(n) / 3

[0067] The specific interpolation filtering process is:

[0068]

[0069] Where K is the frame length of the burst data sequence and K = K p +K d .

[0070] S5. Perform differential frame synchronization operation on the data sequence y(k) recovered in step S4: select a data sequence of appropriate length, perform differential operation on the preceding and following symbols in the sequence, and similarly select a pilot sequence of the same length, perform correlation accumulation operation on the result after the differential operation, and perform peak detection on the accumulated result;

[0071] The differential frame synchronization related calculation method is:

[0072]

[0073] Among them, p h It is a part of the pilot symbols in the physical layer frame, with a length of K sp ; sign(·) is the sign function, which is used to get the sign bit of the symbol in the data sequence; real(·) represents the real part of the complex signal; imag(·) represents the imaginary part of the complex signal; K sp is the pilot length used for correlation calculation, which must be smaller than the pilot length in the burst transmission sequence; (·) * Represents the conjugate calculation operation.

[0074] S6. Determine whether the frame synchronization process is completed based on the peak detection result, and determine whether the peak value is greater than a preset threshold; if satisfied, complete timing synchronization and frame synchronization; if not satisfied, repeat steps S2 to S6 until synchronization.

[0075] Example 2

[0076] like Figures 1-2 As shown, a fast differential frame synchronization method under a burst transmission system includes the following steps:

[0077] In this embodiment, the method is used to solve the frame synchronization problem under the satellite communication burst transmission system. Specifically, when the original bit stream undergoes 1 / 2LDPC encoding and QPSK modulation, the physical frame contains 388 pilot symbols and 4096 data symbols, the selected observation interval length is 128 symbols, and the selected pilot length for correlation calculation is 100 symbols.

[0078] The specific implementation method of this method is as follows:

[0079] S1. The receiver performs matched filtering and downsampling on the received burst transmission sequence, reducing the sampling rate of the burst sequence to 4 times the symbol rate. The pilot content is a fixed 0 / 1 sequence with good autocorrelation characteristics. The length of the data part is K. d , the data content is a random 0 / 1 sequence;

[0080] S2. Starting from the first received sampling point, 512 sampling points are selected in sequence for accumulation operation. Each group of sampling points is updated in a fixed interval sliding manner. The sliding interval is 4 sampling points. For example, the first group of x(1) sampling points is x(0), x(1), ..., x(511), then the second group of x(2) sampling points is x(4), x(5), ..., x(515), and so on. The timing error is calculated using the following formula:

[0081]

[0082] Get n timing error results in sequence

[0083] S3. Based on the timing error calculation result, the integer parameter m(n) and the fractional parameter μ(n) are calculated in sequence using the following formulas:

[0084]

[0085] After calculation, n integer parameters and fractional parameters are obtained in turn, namely m(1), m(2),…, m(n) and μ(1), μ(2),…, μ(n).

[0086] S4. Use cubic interpolation filter to calibrate timing error and recover the correct data sequence:

[0087] The cubic (i=-2,-1,0,1) interpolation filter coefficients are:

[0088] c-2 (μ)=μ(n) 3 / 6-μ(n) / 6

[0089] c -1 (μ)=-μ(n) 3 / 2+μ(n) 2 / 2+μ(n)

[0090] c0(μ)=μ(n) 3 / 2-μ(n) 2 -μ(n) / 2+1

[0091] c1(μ)=-μ(n) 3 / 6+μ(n) 2 / 2-μ(n) / 3

[0092] The filtered data sequence is:

[0093]

[0094] Select the best filtered data sequence among the four sampling points.

[0095] S5. Perform differential frame synchronization operation using the data sequence y(n) recovered in step S4:

[0096] The differential frame synchronization related calculation method is:

[0097]

[0098] Among them, p h are 100 consecutive pilot symbols in the physical layer frame. In this embodiment, the 128th to 227th pilot symbols are selected as the local pilot sequence for cross-correlation.

[0099] S6. The differential frame synchronization result in step S5 is compared with the preset thresholds in turn. When |R(u)|>R T At this time, n is the starting position of the data frame header, which is marked as N F , the peak detection diagram is as follows Figure 3 As shown, the frame header is located at position 128.

[0100] S7, when n = 129 ~ 4357 (symbol), continue to perform steps S2, S3 and S4 to obtain the output signal y (n); when n = 4358 ~ 4484, use the timing error result The integer parameter m(4357) and the fractional parameter μ(4357) are interpolated by cubic interpolation to obtain the output signal y(n).

[0101] In order to test the false capture probability of the algorithm under the demodulation threshold, set Eb The value of / N0 is 5dB, the frequency deviation is set to 20KHz, the phase deviation is set to π / 4, and the number of simulations is 1000. The differential frame synchronization correlation peaks obtained by the algorithm are compared to see if there is false capture or missed capture. Figure 4 ,In 1000 simulation experiments, the correlation peak value has been maintained at around 20, ,the frame synchronization position captured each time is fixed at 127, and the ,number of false captures and missed captures are both 0, ,which meets the capture probability index of more than 99%.

[0102] Assume that the burst transmission signal E b When / N0 is 2 to 15, the receiver uses the output signal obtained by the method of the present invention to compare with the original signal output by the host computer, and the bit error rate is calculated. The result shows that E b The corresponding relationship between / N0 and bit error rate conforms to the theoretical bit error curve of QPSK modulation and demodulation. The results are as follows: Figure 5 As shown, it is illustrated that the signal recovered by the method of the present invention is effective and can reach the theoretical demodulation threshold of 1 / 2 LDPC coding.

[0103] Assume that the burst transmission signal E b When / N0 is set to 4, the symbol transmission rate is 200Ksys, and the frequency deviation is set to -30KHz to 50KHz, the output signal obtained by the receiver using the method of the present invention is compared with the original signal output by the host computer, and the bit error rate is counted. The results show that when the absolute value of the frequency deviation set by the system is greater than 30KHz, the bit error rate of the received signal increases. The results are as follows: Figure 6 As shown, this method has a limited range for the frequency deviation carried by the burst transmission signal. When implemented specifically, simulation verification needs to be performed in combination with the symbol transmission rate and different frequency deviation settings.

[0104] When the differential frame synchronization method is used to capture and track dynamic signals with different initial phase deviations, 100 simulation tests are performed under a demodulation threshold of 5 dB. Figure 7 From the simulation results, it can be seen that the dynamics of traversing -π to π phase offsets will not affect the demodulation performance of the receiver. Figure 7 The simulation results show that the algorithm works well in harsh channel environments with large dynamics and low demodulation thresholds, and has the ability to directly capture and track signals.

[0105] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A fast differential frame synchronization method under a burst transmission system, characterized in that: The following steps are involved: S1. The receiver performs matched filtering and downsampling on the received burst transmission sequence, and reduces the sampling rate to N times the symbol rate to obtain a downsampling sequence, wherein the downsampling sequence includes a pilot sequence and a data sequence, the total length of the downsampling sequence is K, and the pilot sequence is of length K. p , a 0 / 1 sequence with fixed content; the data sequence is of length K d , 0 / 1 sequence with random content; A down-sampling sequence x(n) is obtained by sliding sampling at a fixed interval, where n is initialized to 1, each of the down-sampling sequences includes NL0 sampling points, the sampling time difference between the first sampling points of two adjacent down-sampling sequences is N, and L0 is the observation interval length; S2. Obtain the timing error of the down-sampling sequence x(n) Wherein, k is the sampling point position index in the data sequence of the down-sampling sequence x(n), T s is the sampling period of the phase-finding operator, arg(·) is the phase-finding operator, and T is the period of the phase-finding operator; S3, based on the timing error The integer parameter m(n) and the fractional parameter μ(n) of the down-sampling sequence x(n) are obtained: S4, using a cubic interpolation filter to calibrate the timing error and recover the correct data sequence y(u); Cubic interpolation filter coefficients c i for: c -2 (μ(n))=μ(n) 3 / 6-μ(n) / 6 c -1 (μ(n))=-μ(n) 3 / 2+μ(n) 2 / 2+μ(n) c0(μ(n))=μ(n) 3 / 2-μ(n) 2 -μ(n) / 2+1 c1(μ(n))=-μ(n) 3 / 6+μ(n) 2 / 2-μ(n) / 3; The cubic interpolation filtering process is: Determine whether the starting position N of the data frame header is obtained F If yes, n=n+1, return to step S2, if no, go to step S5; S5, selecting a data sequence y(u) of a specified length and performing a differential frame synchronization operation on the preceding and following symbols in the sequence to obtain a differential frame synchronization result R(u); Among them, p h is part of the pilot symbol in the physical layer frame, with a length of K sp , K sp ≤K p ;p u is a pilot sequence of the same length as the data sequence y(u) of the specified length; sign(·) is a sign function used to obtain the sign bit of the symbol in the data sequence; real(·) is the real part of the complex signal; imag(·) is the imaginary part of the complex signal; (·) * It is a conjugate calculation operation; S6, compare the differential frame synchronization result R(u) with the threshold R T For comparison, when |R(u)|>R T When u is the starting position of the data frame header, marked as N F ; n=n+1, return to step S2; Until all burst transmission sequence processing is completed, a fast differential frame synchronization method under a burst transmission system is completed.

2. The fast differential frame synchronization method under a burst transmission system according to claim 1, characterized in that: In step S1, N is 4, and the sampling rate is 4 times; the length of the observation interval is related to the timing error estimation accuracy.

3. The fast differential frame synchronization method under a burst transmission system according to claim 1, characterized in that: When the number of sampling points in the down-sampling sequence x(n) in step S1 is less than NL0, the timing error is used The integer parameter m(n-1) and the fractional parameter μ(n-1) perform cubic interpolation filtering in step S4.

4. The fast differential frame synchronization method under a burst transmission system according to claim 1, characterized in that: If communication interruption occurs during the communication process, the sampling in step S1 is performed again.

5. The fast differential frame synchronization method under a burst transmission system according to claim 1, characterized in that: L0≤K-K p 。 6. The fast differential frame synchronization method under a burst transmission system according to claim 1, characterized in that: In step S1, the receiver simultaneously receives synchronization information, the synchronization information includes: sending time slot number, receiving time slot number, TOD, frequency hopping time difference and frequency hopping rule; the frequency hopping rule includes: frequency hopping speed, frequency hopping block size and number of frequency hopping blocks per unit time slot.

7. The fast differential frame synchronization method under a burst transmission system according to claim 1, characterized in that: In step S6, the threshold R T It is set according to the demodulation threshold, missed alarm probability and false alarm probability. If normalization is performed, the threshold R T Not greater than the length of the pilot part selected in step S5.

8. The fast differential frame synchronization method under a burst transmission system according to claim 1, characterized in that: The fast differential frame synchronization method uses a timing synchronization device based on a time slot frequency hopping network, and the timing synchronization device includes a transmitter processor and a receiver processor.

9. The fast differential frame synchronization method under a burst transmission system according to claim 8, characterized in that: The sending processor and the receiving processor are both FPGAs.

Images