A composite differential frame synchronization method for burst communication under high dynamic conditions

Abstract

The present invention relates to a composite differential frame synchronization algorithm for burst communication under high dynamic conditions. The algorithm comprises the following steps: Step 1: differentially processing the received signal. In high dynamic scenarios, differential processing is required to reduce the influence of frequency deviation and obtain a differential sequence; Step 2: processing multiple groups of differential sequences in parallel, repeating Step 1, wherein each differential process differs in that the delay period of the differential is different, i.e., obtaining multiple groups of differential sequences with different delay periods; then sliding the phase of each group of differential sequences, and using the maximum value as the cross-correlation value of the differential sequence after passing through a filter, etc. The method of the present invention has the following advantages: it is easy to implement in hardware, has low complexity, can quickly obtain frame synchronization results, and can then quickly demodulate the signal to obtain information. It is suitable for burst signals, has real-time performance, and has wide application value in military battlefields.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communication in high dynamic scenarios, and specifically refers to a frame synchronization method for burst signals under the conditions of low signal-to-noise ratio and high dynamics coexisting. Background Art

[0002] Burst communications, widely used in military and satellite applications, are characterized by the uncertainty of signal arrival, duration, and end time. Parameter estimation, modulation recognition, and demodulation often assume that the signal has been received. However, detecting the presence of the target signal in the received data and determining its start and end times are crucial tasks in burst signal processing.

[0003] Frame synchronization is a critical operation in burst communication systems, marking the boundaries of information data within a data stream. To achieve frame synchronization at the receiving end, a preamble is added to the header of each burst signal. The preamble typically consists of two segments: a bit timing recovery (BTR) word and a unique word (UW). These are used for bit timing recovery and frame synchronization, respectively. However, frame synchronization for burst signals faces many challenges. First, bursts can be very short, or the interval between two consecutive bursts can be very small. These situations complicate burst signal acquisition and frame synchronization. Second, due to the high dynamics of transmission at both the transmitter and receiver, significant Doppler shift is inevitably present in the received signal, causing signal distortion and potentially leading to frame synchronization failure.

[0004] Correlation estimation and maximum likelihood estimation are two key concepts for solving this problem. While maximum likelihood estimation outperforms correlation estimation, it is computationally intensive and complex, making it unsuitable for burst communications. Correlation estimation is popular due to its simplicity and ease of hardware implementation. It exploits the uncorrelated nature of noise, improving the signal-to-noise ratio (SNR). However, correlation rules are highly sensitive to frequency and phase. When differential processing is performed on the received signal to eliminate the effects of large frequency offsets, the differential operation degrades the algorithm's SNR. Consequently, existing correlation rules cannot simultaneously address high-dynamic and low-SNR environments.

[0005] In the published patent literature, for example, Chinese invention patent application No. 202110250436.6 discloses a frame synchronization method and a frame synchronization module. For the received data obtained at each moment, the method includes: using differential correlation calculation on the received data at the current moment to obtain the differential correlation sequence corresponding to the current moment; obtaining a local correlation sequence of differential related information representing the SOF sequence; performing phase judgment on each data in the differential correlation sequence corresponding to the current moment to obtain a symbol sequence; performing XOR processing on the local correlation sequence and the symbol sequence to obtain a decision weight; using the differential correlation sequence corresponding to the current moment and the local correlation sequence to obtain a correlation value; multiplying the decision weight correlation value to obtain a decision value; comparing the decision value with a preset threshold to obtain the capture result of the frame start position at the current moment. The present invention first performs differential correlation calculation and hard decision on the received data, and uses the accumulated result obtained based on the simplified differential posterior accumulation algorithm as the weighted value of the correlation value, which can improve the anti-frequency deviation characteristics and the probability of detecting the frame start position.

[0006] For another example, Chinese invention patent application 202110250441 discloses a frame synchronization method, device, electronic device and storage medium. For the received data obtained at each moment, the method includes: using differential correlation calculation on the received data at the current moment to obtain the differential correlation sequence corresponding to the current moment; obtaining a local correlation sequence of differential related information representing the SOF sequence; performing phase judgment on each data in the differential correlation sequence corresponding to the current moment to obtain a symbol sequence; performing XOR processing on the local correlation sequence and the symbol sequence to obtain a decision value; comparing the decision value with a preset threshold to obtain a capture result of the frame start position at the current moment. The present invention first performs differential correlation calculation on the received data, which can improve the algorithm's anti-frequency deviation characteristics, and thus can increase the detection probability of the frame start position in the presence of frequency deviation. Only using the sign bit of the received data and the local correlation sequence to correlate to obtain the decision value can effectively reduce the computational complexity, reduce resource occupancy, and facilitate hardware implementation of the algorithm.

[0007] For another example, Chinese invention patent application No. 200910148394.4 discloses a frame synchronization method for an OFDM system, comprising: (a) performing a segmented cross-correlation operation on a received signal sequence and a local synchronization signal sequence, and normalizing the result to obtain a normalized segmented cross-correlation result corr_cross(k); (b) performing an autocorrelation operation on the normalized segmented cross-correlation result corr_cross(k), and normalizing the result to obtain a normalized autocorrelation result corr_auto(k); (c) performing a differential operation on the normalized autocorrelation result corr_auto(k) to obtain a differential operation result diff_value(k), and taking the first peak point where diff_value(k) exceeds a preset threshold as the starting position of the frame; and (d) adjusting the position of the received signal sequence according to the frame starting position determined in step (c), and outputting a synchronized received signal. The frame synchronization method of the present invention can not only operate in a multipath channel, but also resist large carrier frequency offsets.

[0008] The above disclosed invention patent applications all have the problem of reducing the signal-to-noise ratio of the algorithm during operation, and cannot overcome the problem of high dynamic and low signal-to-noise ratio environments. Summary of the Invention

[0009] In view of the deficiencies of the prior art, the present invention provides a composite differential frame synchronization algorithm for burst communication under high dynamic conditions.

[0010] The composite differential frame synchronization algorithm for burst communication under high dynamic conditions comprises the following steps:

[0011] Step 1: Perform differential processing on the received signal. In high-dynamic scenarios, the impact of frequency offset on frame synchronization cannot be ignored. Therefore, differential processing is required to reduce the impact of frequency offset and obtain a differential sequence.

[0012] Step 2: Process multiple groups of differential sequences in parallel. Repeat step 1. The difference between each differential is the delay period of the differential. That is, multiple groups of differential sequences with different delay periods are obtained. Then, each group of differential sequences is phase-sliding, and the phases are passed through the corresponding filters. The maximum value is used as the cross-correlation value of the differential sequence after passing through the filter.

[0013] Step 3: Analyze all cross-correlation values ​​to identify factors that affect frame synchronization success. Design a decision threshold based on the cross-correlation value and noise power. Then, perform a weighted average of all cross-correlation values ​​and compare the result with the threshold to determine whether a signal has arrived.

[0014] Step 4: Estimate the signal frame position and select the maximum value in the part above the threshold. The position corresponding to this value is the frame synchronization position.

[0015] Furthermore, in step 1, the received signal is differentially processed. In a high-dynamic scenario, the impact of frequency offset on frame synchronization cannot be ignored. It is necessary to perform differential processing on the received signal to weaken the impact of frequency offset and obtain a differential sequence:

[0016] Taking the GMSK signal as the specific modulation signal for burst communication, assuming that the UW part used for frame synchronization in the transmitted signal is s(t), it can be expressed as the following formula (1):

[0017]

[0018] Phase information As shown in formula (2)

[0019]

[0020] Where: I n represents the signal symbol, h is the modulation index of the GMSK signal, T b is the symbol period, q(t) is the integral function of the impulse response of the Gaussian filter, t is the time, k is an integer whose value range is (-∞, n], and the received signal r(t-τ) is expressed as the following formula (3):

[0021]

[0022] Where: f d (t) is the Doppler frequency offset, which consists of fixed frequency offset and variable frequency offset. The fixed frequency offset is mainly caused by the speed difference between the transmitting antenna and the receiving antenna, while the variable frequency offset is caused by the change in signal speed during transmission. θ is the phase offset, which mainly comes from the difference in the initial carrier phase between the transmitting end and the receiving end. n(t) is a Gaussian process with a mean of 0 and a bilateral power spectral density of N0 / 2. τ represents the transmission delay. The receiving end cannot directly identify the signal starting point and the optimal sampling time. Frame synchronization requires estimating the value of τ.

[0023] Taking the same length r(t) as the duration of the impulse response, inspired by the differential idea, the signal is differentiated to obtain the following formula (4):

[0024]

[0025] Where: Δθ(t;k) = f d (t)tf d (t-kT b )(t-kT b )……(5),

[0026]

[0027] Furthermore, in step 2, multiple groups of differential sequences are processed in parallel, and step 1 is repeated. The difference between each differential is that the delay period of the differential is different, that is, multiple groups of differential sequences with different delay periods are obtained. Then, each group of differential sequences is subjected to phase sliding, and the phases are passed through the corresponding filters. The maximum value among them is used as the cross-correlation value of the differential sequence after passing through the filter:

[0028] Since UW is known at the receiving end, the impulse response can be designed to be H at the receiving end. k (t) filter, H k (t) is expressed as the following formula (7):

[0029]

[0030] The above formula (7) represents the signal difference after a delay of k cycles, where:

[0031]

[0032] In formula (8), AT b Represents the length of the synchronization header;

[0033] Knowing that there is still residual frequency offset in Δθ(t; k), we use the sliding phase method to calculate R k (t) performs sliding phase processing, which is expressed as the following formula (9):

[0034] R k (t-τ|α)=R k (t-τ)e jα ……(9),

[0035] Among them: α slides in [0, 2π], R k (t-τ|α) after impulse response is H k (t) filter, taking the maximum value among them to realize the cross-correlation operation between the signal and the local sequence, which is expressed as follows (10):

[0036]

[0037] The sliding phase selects the maximum mutual value corresponding to the change of α value, and the result is expressed as the following formula (11):

[0038] Γ k (τ) = max{Γ k (τ|α)}……(11).

[0039] Furthermore, in step 3, all cross-correlation values ​​are jointly analyzed to analyze the factors that affect the success of frame synchronization. A decision threshold is designed based on the cross-correlation value and noise power. A weighted average of all cross-correlation values ​​is taken and the result is compared with the threshold to determine whether a signal has arrived:

[0040] Define all operations in step 1 and step 2 as a differential cross-correlation box after a delay of K cycles, which is defined as KDSC operation, where K represents the length of the delayed cycle, D represents the differential step, S represents the sliding phase, and C represents the cross-correlation operation. The signal processing procedure at the receiving end is as follows;

[0041] It is deduced that the Γ obtained after kDSC operation is k Since (τ) is related to k, multiple different kDSCs are used for parallel processing and then combined to estimate the time delay τ to see if the lost noise performance can be compensated. After parallel processing, the signal is weighted and summed. When M different kDSCs are running in parallel, the decision term becomes the following equation (12):

[0042]

[0043] Where: k = 1, 2, ..., M, the specific analysis is as follows:

[0044] First, we analyze the k The signal term in (τ) is expressed as the following formula (13):

[0045]

[0046] in: The signal term in its weighted average is expressed as follows (15):

[0047]

[0048] Since GMSK has the characteristic of constant envelope, when the received signal and the filter are completely matched, the following equation (16) is obtained:

[0049]

[0050] in: Refers to the delay estimated using the frame synchronization method;

[0051] When the match is not complete, that is, not aligned, Υ k (τ) changes with the change of k and τ. After weighted summation, the overall trend of Υ(τ) is downward. By performing multiple kDSC (k = 1, 2, ..., M) and weighted averaging, the autocorrelation characteristics of the signal can be improved and the interference term can be effectively reduced. The value of

[0052] At the same time, Γ k The noise term in (τ) is expressed as follows (17):

[0053]

[0054] As mentioned above, frequency deviation, noise, and signal are independent of each other, so we have the following equation (18):

[0055]

[0056] in: and P s(t) are the power of the frequency offset part and the signal term part in the received signal respectively.

[0057] Derived noise power is the following formula (19):

[0058]

[0059] The noise term after weighted averaging is as follows (20):

[0060]

[0061] As we know above, due to the difference of k, the noise term Ω k The incomplete consistency of (τ)(k=1, 2, ..., M) is obtained based on formula (19):

[0062]

[0063] Where: P Ω(τ) The specific mathematical model is as follows (22):

[0064]

[0065] represents the noise power of each group of differential sequences, that is, the following formula (23):

[0066]

[0067] In formula (23), P ζ(τ) represents the cross-correlation value between the noise in each differential sequence, that is, the following formula (24):

[0068]

[0069] in: Using the following formula (26):

[0070]

[0071] The following formula (27) is obtained:

[0072]

[0073] As mentioned above, after performing the composite differential cross-correlation frame synchronization algorithm, the noise power can be reduced and the signal-to-noise ratio can be improved. The threshold is defined as follows (28):

[0074]

[0075] Among them: μ1, η, μ2 are constants that balance the false alarm rate and false alarm rate to meet various indicators in different projects.

[0076] Furthermore, in step 4, the signal frame position is estimated, and the maximum value is selected in the part above the threshold. The position corresponding to the maximum value is the frame synchronization position:

[0077] when:

[0078] V th ≤[Γ(τ1),Γ(τ2)],

[0079] It is assumed that the signal reaches the receiving end, and the delay τ is roughly estimated at this time, as shown in the following formula (29):

[0080]

[0081] The present invention has the following superior technical effects:

[0082] The composite differential frame synchronization method for burst communications under high dynamics, described in the present invention, significantly reduces the impact of Doppler frequency deviation through differential operations. The differential sequence is cross-correlated with the local sequence, and the characteristic that the signal and noise terms in the cross-correlation value vary with the differential delay period is analyzed and utilized. Differential cross-correlations are performed on signals with different delays to obtain signal terms and noise power with different characteristics. This characteristic is exploited through weighted summation, significantly reducing the signal-to-noise ratio degradation caused by differentials and simultaneously reducing interference from signal terms at asynchronous locations. Therefore, the composite differential frame synchronization method solves the problem of signal frame synchronization in scenarios with both high dynamics and low signal-to-noise ratio. Compared with traditional differential cross-correlation methods, the method has stronger noise immunity and can compensate for the signal-to-noise ratio loss caused by differential operations.

[0083] The composite differential frame synchronization method for burst communication under high dynamic conditions described in the present invention is easy to implement in hardware, and can quickly obtain frame synchronization results and then quickly demodulate signals to obtain information. It is suitable for burst signals, has real-time performance, and has wide application value in military battlefields. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Figure 1 This is a flow chart of the composite differential frame synchronization method for burst communication under high dynamic conditions according to the present invention.

[0085] Figure 2The present invention is a schematic diagram of the preamble structure of the burst signal frame header of the composite differential frame synchronization method for burst communication under high dynamic conditions.

[0086] Figure 3 This is a flow chart of signal processing at the receiving end of the composite differential frame synchronization method for burst communication under high dynamic conditions according to the present invention. DETAILED DESCRIPTION

[0087] In order to better understand the composite differential frame synchronization algorithm for burst communication under high dynamic conditions described in the present invention, the technical solutions in the embodiments of the present invention are clearly and completely described below in conjunction with the accompanying drawings.

[0088] like Figure 1 As shown, the composite differential frame synchronization algorithm method for burst communication under high dynamic conditions of the present invention includes the following steps:

[0089] In step 1, differential processing is performed on the received signal. In a high-dynamic scenario, the impact of frequency offset on frame synchronization cannot be ignored. Differential processing is required on the received signal to reduce the impact of frequency offset and obtain a differential sequence:

[0090] Taking the GMSK signal as the specific modulation signal for burst communication, assuming that the UW part used for frame synchronization in the transmitted signal is s(t), it can be expressed as the following formula (1):

[0091]

[0092] Phase information As shown in formula (2)

[0093]

[0094] Where: I n represents the signal symbol, h is the modulation index of the GMSK signal, T b is the symbol period, q(t) is the integral function of the impulse response of the Gaussian filter, t is the time, k is an integer whose value range is (-∞, n], and the received signal r(t-τ) is expressed as the following formula (3):

[0095]

[0096] Where: f d(t) is the Doppler frequency offset, which consists of fixed frequency offset and variable frequency offset. The fixed frequency offset is mainly caused by the speed difference between the transmitting antenna and the receiving antenna, while the variable frequency offset is caused by the change in signal speed during transmission. θ is the phase offset, which mainly comes from the difference in the initial carrier phase between the transmitting end and the receiving end. n(t) is a Gaussian process with a mean of 0 and a bilateral power spectral density of N0 / 2. τ represents the transmission delay. The receiving end cannot directly identify the signal starting point and the optimal sampling time. Frame synchronization requires estimating the value of τ.

[0097] Taking the same length r(t) as the duration of the impulse response, inspired by the differential idea, the signal is differentiated to obtain the following formula (4):

[0098]

[0099] Where: Δθ(t;k) = f d (t)tf d (t-kT b )(t-kT b )……(5),

[0100]

[0101] Step 2: Process multiple groups of differential sequences in parallel. Repeat step 1. The difference between each differential is that the delay period of the differential is different. That is, multiple groups of differential sequences with different delay periods are obtained. Then, each group of differential sequences is phase-sliding, and the phases are passed through the corresponding filters. The maximum value is used as the cross-correlation value of the differential sequence after passing through the filter:

[0102] Since UW is known at the receiving end, the impulse response can be designed to be H at the receiving end. k (t) filter, H k (t) is expressed as the following formula (7):

[0103]

[0104] The above formula (7) represents the signal difference after a delay of k cycles, where:

[0105]

[0106] In formula (8), AT b Represents the length of the synchronization header;

[0107] Knowing that there is still residual frequency offset in Δθ(t; k), we use the sliding phase method to calculate R k (t) performs sliding phase processing, which is expressed as the following formula (9):

[0108] R k (t-τ|α)=Rk (t-τ)e jα ……(9),

[0109] Among them: α slides in [0, 2π], R k (t-τ|α) after impulse response is H k (t) filter, taking the maximum value among them to realize the cross-correlation operation between the signal and the local sequence, which is expressed as follows (10):

[0110]

[0111] The sliding phase selects the maximum mutual value corresponding to the change of α value, and the result is expressed as the following formula (11):

[0112] Γ k (τ) = max{Γ k (τ|α)}……(11);

[0113] Step 3: Jointly analyze all cross-correlation values ​​to identify factors that affect frame synchronization success. Design a decision threshold based on the cross-correlation value and noise power. Perform a weighted average of all cross-correlation values ​​and compare the result with the threshold to determine whether a signal has arrived.

[0114] All operations in step 1 and step 2 are defined as boxes of differential cross-correlation after a delay of K cycles, which is defined as KDSC operation, where K represents the length of the delayed cycle, D represents the differential step, S represents the sliding phase, and C represents the cross-correlation operation. Then the signal processing flow at the receiving end is as follows: Figure 3 As shown;

[0115] It is deduced that the Γ obtained after kDSC operation is k (τ) is related to k. Multiple different kDSCs are used for parallel processing and then combined to estimate the time delay τ to see if the lost noise performance can be compensated. After parallel processing, the signal is weighted and summed. When M different kDSCs are running in parallel, the decision term becomes the following formula (12):

[0116]

[0117] Where: k = 1, 2, ..., M, the specific analysis is as follows:

[0118] First, analyze the k The signal term in (τ) is expressed as the following formula (13):

[0119]

[0120] in: The signal term in its weighted average is expressed as follows (15):

[0121]

[0122] Since GMSK has the characteristic of constant envelope, when the received signal and the filter are completely matched, the following equation (16) is obtained:

[0123]

[0124] in: Refers to the delay estimated using the frame synchronization method;

[0125] When the match is not complete, that is, not aligned, Υ k (τ) changes with the change of k and τ. After weighted summation, the overall trend of Υ(τ) is downward. By performing multiple kDSC (k = 1, 2, ..., M) and weighted averaging, the autocorrelation characteristics of the signal can be improved and the interference term can be effectively reduced. The value of

[0126] At the same time, Γ k The noise term in (τ) is expressed as follows (17):

[0127]

[0128] As mentioned above, frequency deviation, noise, and signal are independent of each other, so we have the following equation (18):

[0129]

[0130] in: and P s(t) are the power of the frequency deviation part and the signal term part in the received signal,

[0131] Derived noise power is the following formula (19):

[0132]

[0133] The noise term after weighted averaging is as follows (20):

[0134]

[0135] As we know above, due to the difference of k, the noise term Ω k The incomplete consistency of (τ)(k=1, 2, ..., M) is obtained based on formula (19):

[0136]

[0137] Where: P Ω(τ) The specific mathematical model is as follows (22):

[0138]

[0139] represents the noise power of each group of differential sequences, that is, the following formula (23):

[0140]

[0141] In formula (23), P ζ(τ) represents the cross-correlation value between the noise in each differential sequence, that is, the following formula (24):

[0142]

[0143] in: Using the following formula (26):

[0144]

[0145] The following formula (27) is obtained:

[0146]

[0147] As mentioned above, after performing the composite differential cross-correlation frame synchronization algorithm, the noise power can be reduced and the signal-to-noise ratio can be improved. The threshold is defined as follows (28):

[0148]

[0149] Among them: μ1, η, μ2 are constants that balance the false alarm rate and false alarm rate to meet various indicators in different projects;

[0150] Furthermore, in step 4, the signal frame position is estimated, and the maximum value is selected in the part above the threshold. The position corresponding to the maximum value is the frame synchronization position:

[0151] when:

[0152] V th ≤[Γ(τ1),Γ(τ2)],

[0153] It is assumed that the signal reaches the receiving end, and the delay τ is roughly estimated at this time, as shown in the following formula (29):

[0154]

[0155] As described above, although the embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to the specific embodiments without departing from the principles and concepts of the present invention, and the scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A composite differential frame synchronization algorithm for burst communication under high dynamic conditions, comprising the following steps: Step 1: Perform differential processing on the received signal. In high-dynamic scenarios, the impact of frequency offset on frame synchronization cannot be ignored. Differential processing is required to weaken the impact of frequency offset and obtain a differential sequence: Perform differential processing on the received signal. In high-dynamic scenarios, the impact of frequency offset on frame synchronization cannot be ignored. Differential processing is required to weaken the impact of frequency offset and obtain a differential sequence: Taking the GMSK signal as the specific modulation signal for burst communication, the UW part used for frame synchronization in the transmitted signal is s(t), which is expressed as the following formula (1): Phase information As shown in formula (2) Where: I n represents the signal symbol, h is the modulation index of the GMSK signal, T b is the symbol period, q(t) is the integral function of the impulse response of the Gaussian filter, t is the time, k is an integer whose value range is (-∞, n], and the received signal r(t-τ) is expressed as the following formula (3): Where: f d (t) is the Doppler frequency offset, which consists of fixed frequency offset and variable frequency offset. The fixed frequency offset is mainly caused by the speed difference between the transmitting antenna and the receiving antenna, while the variable frequency offset is caused by the change in signal speed during transmission. θ is the phase offset, which mainly comes from the difference in the initial carrier phase between the transmitting end and the receiving end. n(t) is a Gaussian process with a mean of 0 and a bilateral power spectral density of N0 / 2. τ represents the transmission delay. The receiving end cannot directly identify the signal starting point and the optimal sampling time. Frame synchronization requires estimating the value of τ. Taking the same length r(t) as the duration of the impulse response, we can differentiate the signal and obtain the following equation (4): where: Δθ(t; k) = f d (t)t - f d (t - kT b )(t - kT b )(5), Step 2: Process multiple groups of differential sequences in parallel and repeat step 1. The difference between each differential is that the delay period of the differential is different, that is, multiple groups of differential sequences with different delay periods are obtained. Then, each group of differential sequences is subjected to sliding phase, and the phases are passed through the corresponding filters. The maximum value is taken as the cross-correlation value of the differential sequence after passing through the filter. Since UW is known at the receiving end, the impulse response H can be designed at the receiving end. k (t) filter, H k (t) is expressed as the following formula (7): The above formula (7) represents the signal difference after a delay of k cycles, where: In formula (8), AT b Represents the length of the synchronization header; Knowing that there is still residual frequency offset in Δθ(t; k), we use the sliding phase method to calculate R k (t) performs sliding phase processing, which is expressed as the following formula (9): R k (t-τ|α)=R k (t-τ)e jα ......(9), Among them: α slides in [0, 2π], R k (t-τ|α) after impulse response is H k (t) filter, taking the maximum value among them, to realize the cross-correlation operation between the signal and the local sequence, which is expressed as follows (10): The sliding phase selects the maximum mutual value corresponding to the change of α value, and the result is expressed as the following formula (11): C k (τ)=max{Γ k (t|a)}......(11); Step 3: Jointly analyze all cross-correlation values ​​to analyze factors that affect successful frame synchronization. Design a decision threshold based on the cross-correlation values ​​and noise power. Then, perform a weighted average of all cross-correlation values. Compare the result with the threshold to determine whether a signal has arrived. Jointly analyze all cross-correlation values ​​to analyze factors that affect successful frame synchronization. Design a decision threshold based on the cross-correlation values ​​and noise power. Then, perform a weighted average of all cross-correlation values. Compare the result with the threshold to determine whether a signal has arrived. Define all operations in step 1 and step 2 as a differential cross-correlation box after a delay of K cycles, which is defined as KDSC operation, where K represents the length of the delayed cycle, D represents the differential step, S represents the sliding phase, and C represents the cross-correlation operation. The signal processing procedure at the receiving end is as follows; Γ obtained after kDSC operation k (τ) is related to k, so multiple different kDSCs are used for parallel processing and then combined to estimate the time delay τ to see if the lost noise performance can be compensated. After parallel processing, the signal is weighted and summed. When M different kDSCs are running in parallel, the decision term becomes the following formula (12): Where: k = 1, 2, ..., M, the specific analysis is as follows: First, we analyze the k The signal term in (τ) is expressed as the following formula (13): in: The signal term in its weighted average is expressed as follows (15): Since GMSK has the characteristic of constant envelope, when the received signal and the filter are completely matched, the following equation (16) is obtained: in: Refers to the delay estimated using the frame synchronization method; When the match is not complete, that is, not aligned, Υ k (τ) changes with the change of k and τ. After weighted summation, the overall trend of Υ(τ) is downward. By performing multiple kDSC (k = 1, 2, ..., M) and weighted averaging, the autocorrelation characteristics of the signal can be improved and the interference term can be effectively reduced. The value of At the same time, Γ k The noise term in (τ) is expressed as follows (17): Frequency deviation, noise, and signal are independent of each other, so we have the following formula (18): in: and P s(t) are the power of the frequency deviation part and the signal term part in the received signal, Noise power is the following formula (19): The noise term after weighted averaging is as follows (20): Due to the difference in k, the noise term Ω k The incomplete consistency of (τ)(k=1, 2, ..., M) is obtained based on formula (19): Where: P Ω(τ) The specific mathematical model is as follows (22): P Ω(τ) =P χ(τ) +P ζ(τ )……(22), P χ(τ) represents the noise power of each group of differential sequences, that is, the following formula (23): In formula (22), P ζ(τ) represents the cross-correlation value between the noise in each differential sequence, that is, the following formula (24): in: Using the following formula (26): The following formula (27) is obtained: As mentioned above, after performing the composite differential cross-correlation frame synchronization algorithm, the noise power can be reduced and the signal-to-noise ratio can be improved. The threshold is defined as follows (28): Among them: μ1, η, μ2 are constants that balance the false alarm rate and false alarm rate; Step 4: Estimate the signal frame position and select the maximum value in the part above the threshold. The position corresponding to this value is the frame synchronization position.

2. According to the composite differential frame synchronization algorithm for burst communication under high dynamic conditions as claimed in claim 1, in step 4, the estimated signal frame position is selected from the portion above the threshold, and the position corresponding to the maximum value is the frame synchronization position: when: V th ≤[Γ(τ1),Γ(τ2)], It is assumed that the signal reaches the receiving end, and the delay τ is roughly estimated at this time, as shown in the following formula (29):

Images