Timing synchronization method based on pseudo-random sequence weighting

Abstract

The invention discloses a timing synchronization method based on pseudo-random sequence weighting. The timing synchronization method comprises the following steps: step 1, generating a repeated constant-amplitude zero autocorrelation sequence; step 2, generating a precursor; step 3, receiving a time domain preamble; step 4, obtaining a correlation peak; step 5, normalizing correlation peak power; step 6, obtaining a synchronization peak; step 7, determining a multipath envelope range in the synchronization peak; 8, calculating power, and carrying out threshold detection; and step 9, finishing timing synchronization, and ending synchronous capture. According to the method disclosed by the invention, the scattered synchronous peak power caused by the multipath effect or the frequency selective fading is concentrated to the maximum peak value in a manner of incoherent combination of the secondary peak power, so that the distinction degree of the peak value and the bottom noise is increased.

Description

Technical Field

[0001] The present invention belongs to the technical field of communication signal processing, and particularly relates to a timing synchronization method based on pseudo-random sequence weighting in the field. The method can be used in communication technical fields such as satellite communication, mobile communication, ionospheric scattering communication and tropospheric scattering communication, and can realize timing synchronization detection for communication systems in channels with large path loss and severe frequency selective fading. Background Art

[0002] Signals transmitted over long distances and multipath wireless channels will cause distortion and distortion, affecting the performance of timing synchronization. Long-distance information transmission has large path loss, and the communication system works under low signal-to-noise ratio, causing the synchronization peak power to decrease, which is not conducive to symbol timing synchronization. Multipath transmission will cause the received signal to be the superposition of multiple transmitted signals with different amplitudes and phases, accompanied by severe frequency selective fading, synchronization peak power dispersion, and poor timing synchronization effect. When the signal is transmitted over long distances and dense multipath wireless channels, the synchronization peak power is low and dispersed, almost submerged in the noise, which is easy to cause false alarms or missed detections, resulting in a sharp drop in the synchronization detection probability performance.

[0003] Existing timing synchronization methods are generally divided into two types: data-assisted and non-data-assisted (blind synchronization). The data-assisted synchronization method uses the known preamble (training sequence) or pilot signal at both ends of the transmitter and receiver for synchronization. It has high accuracy and generally small amount of calculation. The disadvantage is that it requires additional training sequences or pilot signals, which causes loss of frequency band resources and power. The non-data-assisted synchronization method directly uses the characteristics of the transmitted data for synchronization. The disadvantage is that statistical features are often extracted through a large amount of received data, and the communication system needs to work under a high signal-to-noise ratio.

[0004] Shanghai Sijiu Microelectronics Co., Ltd. disclosed a method for generating and receiving synchronization signals in its patent application "A method for generating and receiving hierarchical synchronization signals and a receiving device" (application number: 2018105528144, application publication number: CN110557349A). In this method, the Kronecker product of multiple layers of synchronization sequences is used to form a synchronization signal. After the synchronization signal is subjected to delayed autocorrelation of the weights at the receiving end each time, the next layer of recursive processing is performed. The disadvantage of this method is that it requires multiple layers of recursive processing, which has high processing complexity; and the method uses a fixed preset threshold for detection, and the timing synchronization effect is poor under a frequency-selective fading channel.

[0005] Sichuan Jiuzhou Electric Group Co., Ltd. disclosed an adaptive threshold signal detection method in its patent application "Adaptive Threshold Controller and Control Method, Direct Spread Signal Detection Method and Circuit" (Application No.: 201510624507.9, Application Publication No.: CN105245247). This method improves the constant false alarm rate (CFAR) detection, corrects the threshold by the sign function ε, and improves the capture performance of the system when the input signal amplitude dynamic range is large. The disadvantage of this method is that in the case of low signal-to-noise ratio, in order to increase the power of the synchronization peak, the preamble or spread spectrum sequence used is very long, and the complexity of the correlation operation is very high.

[0006] Most of the existing timing synchronization methods correlate the received data with the local sequence in the time domain, obtain the synchronization peak for peak detection, and then use threshold detection to complete the timing synchronization. However, under low signal-to-noise ratio, the preamble used is generally long, and the amount of calculation and complexity are high when correlating at the receiving end; under dense multipath or frequency selective fading channels, the synchronization peak power is dispersed and may be submerged in the background noise, which is easy to cause misjudgment or missed detection, resulting in poor timing synchronization detection performance. Summary of the invention

[0007] The technical problem to be solved by the present invention is to provide a timing synchronization method based on pseudo-random sequence weighting, which can improve the synchronization detection probability under low signal-to-noise ratio, dense multipath or frequency selective fading channels.

[0008] The present invention adopts the following technical solution:

[0009] A timing synchronization method based on pseudo-random sequence weighting, the improvement of which is that it comprises the following steps:

[0010] Step 1, generate a repeated constant amplitude zero autocorrelation sequence:

[0011] Generate a constant amplitude zero autocorrelation sequence c of fixed length N n , repeated as a repeating unit, the number of repetitions is a pseudo-random sequence p The length P;

[0012] Generate a constant amplitude zero autocorrelation sequence c of fixed length N n The formula is as follows:

[0013]

[0014] In the above formula, n = 0, 1, ..., N-1, Q is an arbitrary integer that is relatively prime to N and greater than 1, and q represents an arbitrary integer;

[0015] Step 2: Use a pseudo-random sequence to weight the repeated constant amplitude zero autocorrelation sequence to generate a preamble:

[0016] Using a pseudo-random sequence a p The repeated constant amplitude zero autocorrelation sequence is weighted, and each repeated unit corresponds to a value in the pseudo-random sequence to generate the leading s m , the length is N×P;

[0017] Step 3, receiving the time domain preamble:

[0018] The preamble passes through a frequency selective fading channel, the impulse response length of the channel is G, G < N, and is interfered by the channel Gaussian white noise. When it reaches the receiving end, the receiver receives the time domain preamble r m ;

[0019] Step 4: Select the leader to be detected and perform correlation operation to obtain the correlation peak:

[0020] The time domain leading r m After the correlation operation is performed by a digital matched filter, the filter coefficient length is N, and it is the conjugate of the inverse order of the constant amplitude zero autocorrelation sequence, and the correlation peak y is obtained. m :

[0021]

[0022] In the above formula, m = 0, 1, ..., M-1, M represents the length of the leading edge, M = N × P, * represents the convolution operation, g n represents the coefficients of the digital matched filter, H represents the conjugate operation, n = 0, 1, ..., N-1;

[0023] Step 5: normalize the correlation peak power to obtain the normalized correlation peak:

[0024] The correlation peak is normalized and calculated according to the power of the received time domain pilot to obtain the normalized correlation peak

[0025]

[0026] In the above formula, |·| 2 Indicates the calculation of the square of the modulus;

[0027] Step 6, perform correlation operation on the normalized correlation peak and the pseudo-random sequence to obtain the synchronization peak:

[0028]

[0029] In the above formula, represents the value of the Np+nth normalized correlation peak, p=0,1,...,P-1;

[0030] Step 7, determine the multipath envelope range in the synchronization peak:

[0031] Search for the location of the synchronization peak, and based on the impulse response length G of the channel, take G / 2-1 multipaths to the left and right of the peak to determine the multipath envelope range in the synchronization peak;

[0032] Step 8: Search for the secondary peak, calculate the power, and incoherently combine the power values ​​to obtain a new synchronization peak as the decision value for threshold detection:

[0033] In the multipath envelope range, search for the secondary peaks on the power delay spectrum to obtain the number of secondary peaks B, 1<B<G, calculate the power of each secondary peak, and then incoherently merge the power values ​​of these secondary peaks into the peak power to obtain the new synchronization peak v′ n As the decision quantity, threshold detection is performed;

[0034] The incoherent combining formula is as follows:

[0035]

[0036] In the above formula, v′ n represents the new synchronization peak, e=0,1,...,B-1, represents the kth synchronization peak e The power value of the secondary peak;

[0037] Step 9: If the decision value is higher than the preset threshold value, the timing synchronization is completed; or the search for the receiving time domain preamble is completed, and the synchronization capture is ended.

[0038] Furthermore, the pseudo-random sequence in step 2 is one of an M sequence, an m sequence, a Gold sequence or a Barker code.

[0039] The beneficial effects of the present invention are:

[0040] The method disclosed in the present invention effectively shortens the length of the local correlation sequence by weighting the pseudo-random sequence, reduces the computational complexity, and can be changed from a P×N order filter to a P+N order compared to the traditional digital matched filter method. Especially under low signal-to-noise ratio, the preamble length is long. If a long sequence is used for correlation, the resource consumption is large, the time is long, and the hardware circuit is complex to implement; the method proposed by the present invention is used for timing synchronization, which can effectively reduce the complexity of the correlation operation and reduce resource consumption.

[0041] The method disclosed in the present invention concentrates the synchronization peak power dispersed by multipath effect or frequency selective fading to the maximum peak value by incoherently combining the secondary peak power, thereby increasing the distinction between the peak value and the background noise. The method proposed in the present invention is used for timing synchronization, which can effectively avoid misjudgment or missed detection of timing synchronization under low signal-to-noise ratio and dense multipath, improve the detection probability of timing synchronization, and achieve better timing synchronization effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic flow diagram of the method of the present invention;

[0043] Figure 2 It is a simulation comparison diagram of the new synchronization peak obtained by the method of the present invention and the synchronization peak of the adaptive threshold synchronization in the prior art;

[0044] Figure 3 It is a simulation comparison diagram of the synchronous detection probability of the present invention and the synchronous detection probability of the adaptive threshold in the prior art. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0046] Embodiment 1. This embodiment discloses a timing synchronization method based on pseudo-random sequence weighting. Different from the traditional method of direct time domain correlation and peak detection, the method of the present invention uses a pseudo-random (PN) sequence to weight the constant amplitude zero autocorrelation (CAZAC) sequence to form a secondary weighted preamble, and performs hierarchical correlation detection at the receiving end, which effectively reduces the complexity of local timing synchronization detection. During peak detection, the multipath secondary peaks are first merged to obtain a new synchronization peak with a more obvious peak, and then adaptive threshold detection is performed. In the case of low signal-to-noise ratio, dense multipath or severe frequency-selective fading channels, the timing synchronization performance of the signal is improved. In addition, the preamble signal has a small envelope fluctuation, which can reduce the cost of the linear power amplifier, enable it to work in the nonlinear power amplifier area, and improve the power utilization of the transmitter. Figure 1 As shown, the specific steps include:

[0047] Step 1, generate a repeated constant amplitude zero autocorrelation sequence:

[0048] Generate a constant amplitude zero autocorrelation sequence c of fixed length N n , repeated as a repeating unit, the number of repetitions is a pseudo-random sequence p The length P;

[0049] Generate a constant amplitude zero autocorrelation sequence c of fixed length N n The formula is as follows:

[0050]

[0051] In the above formula, n=0,1,...,N-1, Q is any integer that is relatively prime to N and greater than 1, q represents any integer; e represents an exponential operation with a natural number as the base;

[0052] Step 2: Use a pseudo-random sequence to weight the repeated constant amplitude zero autocorrelation sequence to generate a preamble:

[0053] Using a pseudo-random sequence a p The repeated constant amplitude zero autocorrelation sequence is weighted, and each repeated unit corresponds to a value in the pseudo-random sequence to generate the leading s m , the length is N×P;

[0054] The pseudo-random sequence is one of M sequence, m sequence, Gold sequence or Barker code.

[0055] Step 3, receiving the time domain preamble:

[0056] The preamble passes through a frequency selective fading channel, the impulse response length of the channel is G, G < N, and is interfered by the channel Gaussian white noise. When it reaches the receiving end, the receiver receives the time domain preamble r m ;

[0057] Step 4: Select the leader to be detected and perform correlation operation to obtain the correlation peak:

[0058] The time domain leading r m After the correlation operation is performed by a digital matched filter, the filter coefficient length is N, and it is the conjugate of the inverse order of the constant amplitude zero autocorrelation sequence, and the correlation peak y is obtained. m :

[0059]

[0060] In the above formula, m = 0, 1, ..., M-1, M represents the length of the preamble, and its value is equal to the product of the length of the constant amplitude zero autocorrelation sequence and the length of the pseudo-random sequence, M = N × P, * represents the convolution operation, g n represents the coefficients of the digital matched filter, H represents the conjugate operation, n = 0, 1, ..., N-1;

[0061] Step 5: normalize the correlation peak power to obtain the normalized correlation peak:

[0062] The correlation peak is normalized and calculated according to the power of the received time domain pilot to obtain the normalized correlation peak

[0063]

[0064] In the above formula, |·| 2 Indicates the calculation of the square of the modulus;

[0065] Step 6, perform correlation operation on the normalized correlation peak and the pseudo-random sequence to obtain the synchronization peak:

[0066]

[0067] In the above formula, represents the value of the Np+nth normalized correlation peak, p=0,1,...,P-1;

[0068] Step 7, determine the multipath envelope range in the synchronization peak:

[0069] Search for the location of the synchronization peak, and based on the impulse response length G of the channel, take G / 2-1 multipaths to the left and right of the peak to determine the multipath envelope range in the synchronization peak;

[0070] Step 8: Search for the secondary peak, calculate the power, and incoherently combine the power values ​​to obtain a new synchronization peak as the decision value for threshold detection:

[0071] In the multipath envelope range, search for the secondary peaks on the power delay spectrum, i.e., the secondary peaks, and obtain the number of secondary peaks B, 1<B<G. Calculate the power of each secondary peak, and then incoherently merge the power values ​​of these secondary peaks into the peak power to obtain the new synchronization peak v′ n As the decision quantity, threshold detection is performed;

[0072] The incoherent combining formula is as follows:

[0073]

[0074] In the above formula, v′ n represents the new synchronization peak, e=0,1,...,B-1, represents the kth synchronization peak e The power value of the secondary peak;

[0075] Step 9: If the decision value is higher than the preset threshold value, the timing synchronization is completed; or the search for the receiving time domain preamble is completed, and the synchronization capture is ended.

[0076] The method of the present invention and the adaptive threshold synchronization method of the prior art are compared and simulated using MATLAB R2019a software. The results are as follows: Figure 2 and Figure 3 shown.

[0077] Figure 2This is a simulation comparison diagram of the new synchronization peak obtained by the method of the present invention and the synchronization peak of the adaptive threshold synchronization of the prior art, wherein the horizontal axis represents the sampling point, and the vertical axis represents the synchronization peak power. In the simulation experiment, the bandwidth of the communication system is 128kHz, the power delay spectrum of the ionospheric scattering channel is double exponentially distributed, the channel type is Rayleigh channel, the measured multipath delay and power are [0,50,100,150,200,250,300,350,400,450]μs and [-10.9,-6.8,-2.2,0,-2,-6,-7.7,-7.8,-8.3,-10.4]dB, respectively, and the length of the channel impulse response is G=58. The CAZAC sequence length used in the preamble is N=509, the pseudo-random sequence is P=13-bit Barker code [1,1,1,1,1,-1,-1,1,-1,1,-1,1], the number of secondary peaks found is B=24, and the signal-to-noise ratio is set to 5 dB.

[0078] The prior art used for simulation comparison with the method of the present invention is an adaptive threshold synchronization method, in which the length of the CAZAC sequence used in the preamble is 6617, and the remaining simulation parameters are the same as those of the method of the present invention. The synchronization peaks obtained by the above two methods are compared.

[0079] Figure 2 The middle dashed line represents the synchronization peak obtained by adaptive threshold synchronization, and the solid line represents the new synchronization peak obtained by the synchronization method of the present invention. It can be seen from the two curves that the new synchronization peak obtained by the synchronization method of the present invention has a higher main path or peak power and a lower secondary peak power than the synchronization peak obtained by adaptive threshold synchronization, which effectively increases the distinction between the peak and the background noise, and is beneficial to the timing synchronization detection in a low signal-to-noise ratio multipath channel.

[0080] Figure 3 It is a simulation comparison diagram of the synchronous detection probability of the present invention and the synchronous detection probability of the adaptive threshold in the prior art, wherein the horizontal axis represents the signal-to-noise ratio in dB, and the vertical axis represents the detection probability. The simulation experimental conditions of the present invention are the same as above, the signal-to-noise ratio range is set to -6:1:10dB, and the simulation is performed 100,000 times at each signal-to-noise ratio, and the synchronous detection probability of the present invention and the synchronous detection probability of the adaptive threshold in the prior art are calculated respectively.

[0081] Figure 3The solid line marked with a circle represents the adaptive threshold synchronization detection probability of the prior art, and the solid line marked with a triangle represents the synchronization detection probability of the present invention. It can be seen from the two curves that the synchronization detection probability performance of the present invention is improved by about 2dB compared with the adaptive threshold synchronization detection probability, that is, when the method of the present invention is used, the detection probability exceeds 99.9% when the signal-to-noise ratio is 3dB, while when the traditional adaptive threshold synchronization method is used, the detection probability can only exceed 99.9% when the signal-to-noise ratio is 5dB. This is because the method of the present invention, by merging the secondary peak power, concentrates the dispersed synchronization peak power caused by the multipath effect on the main path, thereby improving the detection performance of timing synchronization under dense multipath.

Claims

1. A timing synchronization method based on pseudo-random sequence weighting, characterized in that: The steps include: Step 1, generate a repeated constant amplitude zero autocorrelation sequence: Generate a constant amplitude zero autocorrelation sequence c of fixed length N n , repeated as a repeating unit, the number of repetitions is a pseudo-random sequence p The length P; Generate a constant amplitude zero autocorrelation sequence c of fixed length N n The formula is as follows: In the above formula, n = 0, 1, ..., N-1, Q is an arbitrary integer that is relatively prime to N and greater than 1, and q represents an arbitrary integer; Step 2: Use a pseudo-random sequence to weight the repeated constant amplitude zero autocorrelation sequence to generate a preamble: Using a pseudo-random sequence a p The repeated constant amplitude zero autocorrelation sequence is weighted, and each repeated unit corresponds to a value in the pseudo-random sequence to generate the leading s m , the length is N×P; Step 3, receiving the time domain preamble: The preamble passes through a frequency selective fading channel, the impulse response length of the channel is G, G < N, and is interfered by the channel Gaussian white noise. When it reaches the receiving end, the receiver receives the time domain preamble r m ; Step 4: Select the leader to be detected and perform correlation operation to obtain the correlation peak: The time domain leading r m After the correlation operation is performed by a digital matched filter, the filter coefficient length is N, and it is the conjugate of the inverse order of the constant amplitude zero autocorrelation sequence, and the correlation peak y is obtained. m : In the above formula, m = 0, 1, ..., M-1, M represents the length of the leading edge, M = N × P, * represents the convolution operation, g n represents the coefficients of the digital matched filter, H represents the conjugate operation, n = 0, 1, ..., N-1; Step 5: normalize the correlation peak power to obtain the normalized correlation peak: The correlation peak is normalized and calculated according to the power of the received time domain pilot to obtain the normalized correlation peak In the above formula, |·| 2 Indicates the calculation of the square of the modulus; Step 6, perform correlation operation on the normalized correlation peak and the pseudo-random sequence to obtain the synchronization peak: In the above formula, represents the value of the Np+nth normalized correlation peak, p=0,1,...,P-1; Step 7, determine the multipath envelope range in the synchronization peak: Search for the location of the synchronization peak, and based on the impulse response length G of the channel, take G / 2-1 multipaths to the left and right of the peak to determine the multipath envelope range in the synchronization peak; Step 8: Search for the secondary peak, calculate the power, and incoherently combine the power values ​​to obtain a new synchronization peak as the decision value for threshold detection: In the multipath envelope range, search for the secondary peaks on the power delay spectrum to obtain the number of secondary peaks B, 1<B<G, calculate the power of each secondary peak, and then incoherently merge the power values ​​of these secondary peaks into the peak power to obtain the new synchronization peak v′ n As the decision quantity, threshold detection is performed; The incoherent combining formula is as follows: In the above formula, v′ n represents the new synchronization peak, e=0,1,...,B-1, represents the kth synchronization peak e The power value of the secondary peak; Step 9: If the decision value is higher than the preset threshold, the timing synchronization is completed; Or after searching for the receiving time domain preamble, the synchronization capture ends.

2. The timing synchronization method based on pseudo-random sequence weighting according to claim 1, characterized in that: The pseudo-random sequence in step 2 is one of an M sequence, an m sequence, a Gold sequence or a Barker code.

Images