Digital receiver system of meteorological satellite

Abstract

The invention discloses a meteorological satellite digital receiver system, and belongs to the technical field of meteorological satellite digital receivers. The meteorological satellite digital receiver system comprises an analog down-conversion module, an analog / digital converter (ADC) module, a digital down-conversion module, a matched filtering module, a positioning synchronization module, a carrier synchronization module, a frame synchronization module, an equalization and automatic gain control (AGC) module, a de-mapping module, an LDPC and BCH decoding module and a baseband data processing module. The timing synchronization module adopts a multi-level Gardner timing synchronization algorithm, the algorithm can process QPSK (Quadrature Phase Shift Keying), 8PSK, 16APSK and 32APSK signals at the same time, and a meteorological satellite digital receiver system can accurately process digital signals.

Description

Technical Field

[0001] The invention belongs to the technical field of meteorological satellites, and in particular relates to a meteorological satellite digital receiver system. Background Art

[0002] Meteorological satellite digital receiver is a vital equipment in the meteorological observation system. It is responsible for receiving and processing digital signals from meteorological satellites, providing key data support for weather forecasting, water resources management, disaster warning and other fields.

[0003] The meteorological satellite digital receiver system includes four modulation modes: QPSK, 8PSK, 16APSK and 32APSK. When receiving signals, a timing error detection algorithm is needed to detect the time difference between the received digital signal and the local clock. The traditional Gardner timing error detection algorithm is usually used. The Gardner timing error detection algorithm is the core key based on the interpolation timing synchronization algorithm. Its advantages are that it has high sensitivity and strong robustness. However, the traditional Gardner timing error detection algorithm is based on BPSK and QPSK, and is not suitable for 8PSK and M-APSK signals of the meteorological satellite digital receiver system. This is because when there is no timing error for single-level signals such as BPSK and QPSK, the middle value of the best sampling points of two adjacent symbols will always return to 0. However, for high-order modulation methods, taking 16QAM modulation as an example, if there is no timing error, the middle point may be 0, -1, +1, -2 and +2. Obviously, if the traditional Gardner error detection formula is used, it will cause error detection errors, thereby causing jitter of the timing clock, and ultimately causing inaccurate digital signals processed by the meteorological satellite digital receiver. Summary of the invention

[0004] The object of the present invention is to provide a meteorological satellite digital receiver system, which can accurately process digital signals and can process QPSK, 8PSK, 16APSK and 32APSK signals at the same time.

[0005] To achieve the above purpose, the present invention provides a meteorological satellite digital receiver system, comprising an analog down-conversion module, an analog / digital converter (ADC) module, a digital down-conversion module, a matched filter module, timing synchronization, carrier synchronization, frame synchronization, equalization and automatic gain control (AGC), demapping, LDPC and BCH decoding, and a baseband data processing module. The radio frequency signal emitted by the satellite is received by an antenna and transmitted to the analog down-conversion module to be converted into an intermediate frequency analog signal, and then the intermediate frequency analog signal is sent to the ADC module to be converted into an intermediate frequency digital signal; the collected intermediate frequency digital signal is multiplied by a local oscillator signal to complete digital down-conversion and output a zero intermediate frequency signal; the zero intermediate frequency signal is filtered by a matched filter module; the filtered signal is processed by timing synchronization, carrier synchronization and frame synchronization, and the data after carrier synchronization will be processed by equalization and AGC. Finally, the decoded data is obtained by demapping, LDPC and BCH decoding in sequence, and is transmitted to the baseband data processing module to obtain a TS data stream, and the data after frame synchronization directly enters the baseband data processing module.

[0006] As a further solution of the present invention: the timing synchronization is used to ensure that the clocks of the receiving end and the transmitting end remain synchronized; the role of the carrier synchronization is to ensure that the local oscillator of the receiving end is synchronized with the carrier frequency and phase of the transmitting end.

[0007] As a further solution of the present invention: timing synchronization adopts a multi-level Gardner timing synchronization algorithm, which includes an interpolation filter, a multi-level Gardner timing error detector, a loop filter and an interpolation controller; input data x(mT s ) enters the interpolation filter, and the interpolation filter outputs the sampled data y(kT i ), and also outputs QPSK, 8PSK, 16APSK, and 32APSK signals, which enter the loop filter after passing through the multi-level Gardner error detection algorithm, and then enter the difference controller after the output signal from the loop filter.

[0008] As a further solution of the present invention: the interpolation filter is used to estimate other sampling points between the sampling points of the discrete signal, so as to achieve the conversion of the data rate; the interpolation filter includes a digital / analog converter, a filter, an analog-to-digital converter, and the data x(mT s ) is successively passed through the digital / analog converter, filter, and analog-to-digital converter to obtain the interpolation output y(kT i ).

[0009] As a further solution of the present invention: the interpolation filter adopts a cubic interpolation algorithm based on a Farrow structure.

[0010] As a further solution of the present invention: the multi-level Gardner timing error detection algorithm includes the following formula:

[0011]

[0012] In the formula [y I (k)+y I (k-1)]·β is the sample point at the middle moment between two symbols, y I (k), y I (k-1) represents the kth and k-1th symbols respectively, β = g T (T / 2) is the impulse response of the pulse shaping filter at time T / 2;

[0013] In the meteorological satellite digital receiver system, the shaping filter uses a root raised cosine filter (RRC), and its impulse response is:

[0014]

[0015] In the formula, α is the roll-off factor of the filter. Substituting t=T / 2 into formula (1-15), β can be derived as:

[0016]

[0017] The specific value of β depends on α. ​​The meteorological satellite digital receiver standard stipulates that the value of α is 0.35, 0.25 and 0.2. Substituting it into formula (1-16), the corresponding β values ​​are 0.619, 0.627 and 0.63 respectively.

[0018] As a further solution of the present invention: the loop filter is used to filter and process the feedback signal of the digital control loop; the loop filter adopts an ideal integral filter.

[0019] As a further solution of the present invention: the function of the interpolation controller is to generate an overflow clock signal, thereby controlling the interpolation filter to perform interpolation at an optimal sampling time.

[0020] Compared with the prior art, the digital receiver system of the meteorological satellite of the present invention adopts a multi-level Gardner timing synchronization algorithm in the timing synchronization module, which includes an interpolation filter, a multi-level Gardner timing error detector, a loop filter and an interpolation controller; the input data x(mT s ) enters the interpolation filter, and the interpolation filter outputs the sampled data y(kT i ), and also outputs QPSK, 8PSK, 16APSK, and 32APSK signals through the multi-level Gardner error detection algorithm and then enters the loop filter. The output signal from the loop filter enters the difference controller. This algorithm can process QPSK, 8PSK, 16APSK, and 32APSK signals at the same time, enabling the meteorological satellite digital receiver system to accurately process digital signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the principle of the meteorological satellite digital receiver system of the present invention.

[0022] Figure 2 It is a schematic diagram of the multi-level Gardner timing synchronization algorithm in the present invention.

[0023] Figure 3 It is a schematic diagram of the model structure of the interpolation filter in the present invention.

[0024] Figure 4 (a) is the cubic interpolator based on the Farrow structure in the present invention.

[0025] Figure 4 (b) is the piecewise parabolic interpolator based on the Farrow structure in the present invention.

[0026] Figure 5 It is a structural diagram of the digital loop filter in the present invention.

[0027] Figure 6 It is the equivalent phase-locked loop in the present invention.

[0028] Figure 7 This is a diagram of the change in the NCO register value in the present invention.

[0029] Figure 8 (a) is the sampling value and interpolation output value of the timing synchronization algorithm in the present invention.

[0030] Figure 8 (b) is a convergence curve diagram of important parameters of the timing synchronization algorithm in the present invention.

[0031] Fig. 9 (a) is the constellation diagram of the QPSK signal before timing synchronization.

[0032] Fig. 9 (b) is the constellation diagram after the QPSK signal is synchronized.

[0033] Fig.10 (a) is the constellation diagram of the 8PSK signal before timing synchronization.

[0034] Fig.10 (b) is the constellation diagram after timing synchronization of the 8PSK signal.

[0035] Fig.11 (a) is the constellation diagram before 16APSK signal timing synchronization.

[0036] Fig.11 (b) is the constellation diagram after the 16APSK signal timing synchronization.

[0037] Fig.12 (a) is the constellation diagram of the 32APSK signal before timing synchronization.

[0038] Fig.12 (b) is the constellation diagram of the 32APSK signal after timing synchronization. Detailed implementation manners

[0039] The present invention will be further described below with reference to the accompanying drawings.

[0040] As Figure 1 shown, a meteorological satellite digital receiver system includes an analog down-conversion module, an analog-to-digital converter (ADC) module, a digital down-conversion module, a matched filtering module, timing synchronization, carrier synchronization, frame synchronization, equalization and automatic gain control (AGC), demapping, LDPC and BCH decoding, and a baseband data processing module. The radio frequency signal emitted by the satellite is received by the antenna and sent to the analog down-conversion module to be converted into an intermediate frequency analog signal, and then the intermediate frequency analog signal is sent into the ADC module to be converted into an intermediate frequency digital signal; the collected intermediate frequency digital signal is multiplied by the local oscillator signal to complete digital down-conversion and output a zero intermediate frequency signal; the zero intermediate frequency signal is filtered by the matched filtering module; the filtered signal is processed through timing synchronization, carrier synchronization and frame synchronization. After carrier synchronization, the data will be processed by equalization and AGC. Finally, the data passes through demapping, LDPC and BCH decoding in sequence to obtain decoded data, which is sent to the baseband data processing module to obtain a TS data stream, and the data after frame synchronization directly enters the baseband data processing module.

[0041] Timing synchronization is used to ensure that the clocks at the receiving end and the sending end are synchronized, so that the received digital signals can be correctly sampled and processed, thereby ensuring the reliability of the data. The function of carrier synchronization is to ensure that the local oscillator at the receiving end is synchronized with the carrier frequency and phase at the sending end. If the carrier synchronization is incorrect, the reliability and accuracy of the received signal will be affected.

[0042] Among them, the multi-level Gardner timing synchronization algorithm is adopted in the timing synchronization process. As Figure 2 shown, the multi-level Gardner timing synchronization algorithm includes an interpolation filter, a multi-level Gardner timing error detector, a loop filter and an interpolation controller; the input data x(mT s ) enters the interpolation filter, and the interpolation filter outputs the sampled data y(kT i ), and also outputs the QPSK, 8PSK, 16APSK, 32APSK signals. After passing through the multi-level Gardner error detection algorithm, they enter the loop filter, and after the signal is output from the loop filter, it enters the difference controller.

[0043] The signal collected by the ADC module at the receiving end is an oversampled signal, that is, one symbol contains multiple sampling points. The function of the interpolation filter is to estimate other sampling points between the sampling points of the discrete signal, so as to achieve the conversion of the data rate. The model of the interpolation filter is as follows: Figure 3 As shown, including digital / analog converter, filter, analog-to-digital converter, data x(mT s ) is successively passed through the digital / analog converter, filter, and analog-to-digital converter to obtain the interpolation output y(kT i );

[0044] Assume that the data collected by the receiver AD is x(mT s ), where T s and T are the sampling clock and symbol period of the receiver respectively. s ) first passes through the digital / analog converter to become a continuous signal, and then passes through the filter h I (t) performs convolution operation, and the output result is:

[0045] y(t)=∑ m x(mT s )h I (t-mT s ) (1-1)

[0046] By i By resampling y(t) at every moment, a new sampling point can be interpolated. The new sampling data y(kT i ) has a period of T i To make the interpolated data match the symbol period, T / T i The value of is an integer. Finally, after resampling, the interpolation filter output is:

[0047] y(kT i )=∑ m x(mT s )h I (kT i -mT s ) (1-2)

[0048] From the above derived formula (1-2), it can be found that if the received sequence x(m) and the impulse response h of the interpolation filter are known, I (t), signal sampling period T s and the interpolation period T i , the above analog interpolation filter can be realized in a fully digital way. So the key is to determine the interpolation period T i The impulse response of the filter h I (t).

[0049] In formula (1-2), m is the index position of the input sequence, and the index of the filter is i=int[kT i / T s ]–m, and then define a new parameter m k = int[kT i / T s ] and μ k =kT i / T s -m k , then the digital interpolation formula is:

[0050]

[0051] Therefore, if you want to adjust the timing by interpolation, you must calculate the control parameter m of the interpolation filter. k With μ k .

[0052] When selecting an interpolation filter, two points should be noted: ⅰ The number of sampling points involved in the interpolation operation should be an even number to ensure that the interpolation filter has a linear phase; ⅱ The interpolation point should be close to the middle of the sampling points, which can solve the delay problem in the interpolation process. Polynomial interpolation can be calculated using the Lagrange interpolation formula of formula (1-4).

[0053]

[0054] In formula (1-4), N 1 =N / 2, N 1 =N / 2-1, when N is an even number, the expression of the interpolation filter can be written as:

[0055] h i [(i+μ k )T s ]=C i (μ k ) (1-5)

[0056] C i is the tap coefficient of the interpolation filter, N 1 With N 2 When N = 2, the interpolation filter is a linear interpolator, C i (μ k ) is:

[0057]

[0058] When N=4, the interpolation filter can be a cubic interpolator and a piecewise parabolic interpolator. Each interpolation calculation requires four sampling points, where the tap coefficient of the cubic interpolator is:

[0059]

[0060] The piecewise parabolic interpolation algorithm uses a four-point quadratic interpolation algorithm to solve the problem that the three-point quadratic interpolation algorithm is not applicable. Its filter tap coefficient is:

[0061]

[0062] Cubic interpolator and piecewise parabolic interpolation are suitable for the nonlinear data distribution of this system. In realizing the above polynomial filter, Farrow structure is a very suitable way for hardware implementation, which has the advantages of simple structure and strong tunability.

[0063] Figure 4 (a) and (b) give the cubic interpolator and piecewise parabolic interpolator based on the Farrow structure respectively.

[0064] from Figure 4 As can be seen from Figure (a), the Farrow structure cubic interpolator contains four vertical operation branches. The output result is the sum of the calculation values ​​of all vertical branches, where D represents the time sampling period T s The delay of each branch f 1 、f 2 、f 3 、f 4 And the output y(kT i ) is calculated as:

[0065]

[0066] from Figure 4 As can be seen from Figure (b), the Farrow structured piecewise parabolic interpolator contains three vertical operation branches, and the output is the calculated values ​​of the first two vertical branches multiplied by the parameter μ k Add the calculated value of the third branch. 1 、f 2 、f 3 And the output y(kT i ) is calculated as:

[0067]

[0068] From the above calculation formula, it can be seen that the two interpolation filters need four consecutive sampling point data x(mT s )、x[(m-1)T s ]、x[(m-2)T s ] and x[(m-3)T s ] and the fractional interval μ k (0≤μ k<1), that is, each interpolation calculation only requires four sampling points.

[0069] The present invention realizes an interpolation filter by adopting a cubic interpolation algorithm based on a Farrow structure.

[0070] The traditional Gardner algorithm is only applicable to BPSK and QPSK signals. This is because when there is no timing error for single-level signals such as BPSK and QPSK, the middle value of the best sampling points of two adjacent symbols will always return to 0. However, for high-order modulation methods, taking 16QAM modulation as an example, if there is no timing error, the middle point may be 0, -1, +1, -2 and +2. Obviously, if the traditional Gardner error detection formula is used, it will cause error detection errors, thereby causing jitter of the timing clock. For the 8PSK and M-APSK signals of meteorological satellite digital receivers, the traditional single-level MGardner algorithm cannot be used. The multi-level Gardner error detection algorithm for high-order QAM can be expressed by formula (1-11).

[0071]

[0072] In formula (1-11), y(k) and y(k-1) represent the kth and k-1th symbols respectively. Considering that the 8PSK and M-APSK signals in the meteorological satellite digital receiver system are multi-level signals and have similarities with the QAM signals.

[0073] Therefore, the present invention applies the multi-level Gardner timing error detection algorithm to M-PSK and M-APSK signals, and the error detection formula thereof will be theoretically derived as follows.

[0074] The baseband signals of M-PSK and M-APSK can be expressed as (1-12):

[0075] r(t)=∑ n c n ·g T (t-kT-τ)+v(t) (1-12)

[0076] Among them, c n is the transmitted symbol sequence, for M-APSK signal, there is c n =R i ·e jθ , R i is the radius of the i-th concentric circle of the APSK signal, θ = 0, 2π / N i ,…,2π(N i -1) / N i The M-PSK signal is a special case of M-APSK. n =R·ejθ τ represents the timing deviation, τ∈[-T / 2,T / 2], g T (t) is the impulse response of the pulse shaping filter, v(t) is Gaussian white noise. The output value of the middle point of two symbols can be expressed by equation (1-13).

[0077]

[0078] In formula (1-13), c n-1 With c n Represent the nth and n-1th symbols respectively, and have the same meaning as y(k) and y(k-1) in formula (1-11), β = g T (T / 2) is the impulse response of the pulse shaping filter at time T / 2. Since the M-APSK signal is also a multi-level signal, the center point zeroing method of the multi-level Gardner algorithm can also be used.

[0079] In addition, this paper also considers the influence of pulse shaping filter on the signal and deduces that the sampling point at the middle moment of two symbols should be [y I (k)+y I (k-1)]·β. Therefore, the multi-level Gardner timing error detection algorithm formula of the M-APSK signal is transformed into formula (1-14):

[0080]

[0081] In the meteorological satellite digital receiver system, the shaping filter uses a root raised cosine filter (RRC), and its impulse response is:

[0082]

[0083] The 0 in the formula is the roll-off factor of the filter. Substituting t=T / 2 into formula (1-15), β can be derived as:

[0084]

[0085] It can be seen that the difference between formula (1-11) and formula (1-14) is that the calculated value of the center point is [y I (k)+y I (k-1)] / 2 becomes [y I (k)+y I (k-1)]·β, because the influence of the root raised cosine filter on the signal waveform is considered here. The specific value of β depends on α. ​​The meteorological satellite digital receiver standard stipulates that the value of α is 0.35, 0.25 and 0.2. Substituting it into formula (1-16), the corresponding β values ​​are calculated to be 0.619, 0.627 and 0.63 respectively.

[0086] The role of the loop filter is to filter and process the feedback signal of the digital control loop so that it can more accurately reflect the state of the controlled object, thereby achieving more accurate control. In the timing synchronization loop, its more important role is to adjust the loop parameters. The role of the interpolation controller is to generate an overflow clock signal to control the interpolation filter to interpolate at the optimal sampling time. The main function of the interpolation controller is to periodically generate an interpolation enable signal to calculate the interpolation base point m of each interpolation point k With fractional interval μ k。

[0087] The present invention uses an ideal integrating filter as a loop filter for timing synchronization.

[0088] The transfer operator and transfer function of the ideal integrating filter are

[0089]

[0090] In order to realize the digital loop filter, it is also necessary to convert from the continuous domain to the discrete domain, that is, from the s domain to the z domain. The most commonly used method is the bilinear transformation method, and its conversion formula is:

[0091]

[0092] Substituting equation (1-18) into the transfer function F(s) of the ideal integrating filter, we can obtain the digital system function of the ideal integrating filter:

[0093]

[0094] Where C 1 With C 2 The values ​​are

[0095]

[0096] According to formula (1-18), the digital circuit structure of the loop filter can be obtained, as follows: Figure 5 shown.

[0097] c in formula (1-20) 1 With c 2 The time constant τ 1 With τ 2 In actual digital implementation, the time constant τ 1 , τ 2 The parameter conversion with digital devices is inconvenient. In order to accurately design the c of the digital loop filter 1 With c 2 , first we need to digitize the system function of the timing synchronization loop. For the timing synchronization loop, it can be equivalent to Figure 5In the second-order type II digital phase-locked loop shown in FIG, the error detector and interpolation controller of the timing synchronization are equivalent to the phase detector and NCO of the phase-locked loop respectively. The loop filter structures of the two are exactly the same, and the interpolation filter has no effect on the loop.

[0098] So for the timing synchronization loop filter parameter c 1 With c 2 The analysis can be transformed into the digital phase-locked loop filter k 1 With k 2 According to the analysis. Figure 6 , it is easy to get the transfer function of the second-order type-2 digital phase-locked loop as:

[0099]

[0100] The output of the phase-locked loop is interfered by a variety of noise sources. Its parameter equivalent noise bandwidth (ENBW) takes into account the influence of these noise sources and gives the comprehensive noise performance of the PLL system. The equivalent noise bandwidth can be derived from formula (1-22).

[0101]

[0102] Substituting H(z) into equation (1-22), the equivalent noise bandwidth of the second-order type II digital phase-locked loop can be obtained as:

[0103]

[0104] Where k 2 <k 1 <<1,B L It can be approximated as:

[0105]

[0106] From the transfer function H(z), it can be seen that the timing synchronization loop is a second-order linear system, and the second-order linear system can generally introduce an undamped oscillation frequency ω n and damping coefficients ξ, ω n and ξ in a digital phase-locked loop are approximated as

[0107]

[0108] In order to obtain the best loop tracking performance, the optimal value of ξ is 0.707, which requires k = 2k 2 , substituting into formula (2-31), we can calculate k 2 , k 1 .

[0109]

[0110] k in formula (1-27) d With k o Represent the gains of the phase detector and NCO respectively. Here, k of the timing loop is set. d =k o =1, in the subsequent simulation and hardware implementation, it is also necessary to ensure that the gains of the two modules are 1. Figure 5 and Figure 6 The loop filter structure in c 1 With c 2 The calculation formula

[0111]

[0112] After the above derivation, the loop filter parameter c 1 With c 2 It becomes only related to the loop equivalent noise bandwidth, which is more convenient for designing loop filters. L T S The size of B determines the loop capture time and noise immunity. L T S When B is larger, the capture time is faster, but the noise immunity is worse. L T S The smaller the value, the slower the capture speed, but the better the noise immunity. In order for the loop to lock normally, B is usually required. L T S <0.1. This paper compromises the capture speed and noise resistance performance and selects B L T S =0.01. Then c is calculated according to formula (1-28): 1 =0.0267, c 2 =0.000356.

[0113] The main function of the interpolation controller is to periodically generate an interpolation enable signal to calculate the interpolation base point m of each interpolation point. k With fractional interval μ k , which is similar to the NCO in the digital phase-locked loop, but the function is slightly different from that of the NCO in the phase-locked loop. The NCO in the phase-locked loop generates the output signal by continuously accumulating the value of the phase register, while the phase register of the interpolation controller changes in a decreasing manner, and its differential equation is:

[0114] η(m+1)=[η(m)-ω(m)]mod1 (1-29)

[0115] In formula (1-29), η(m) and ω(m) represent the phase register value and frequency control word of the mth clock NCO respectively. The interpolation period of the interpolation filter and the working clock period of the interpolation controller are T i With TS , when the timing synchronization loop is stable, the interpolation period of the interpolation filter is also stable, and the value of ω(m) is also approximately a constant. At this time, the interpolation controller will generate an overflow signal every 1 / ω(m) Ts period, from which ω(m) is derived as:

[0116] ω(m)≈T s / T i (1-30)

[0117] Formula (1-30) gives the estimated value of ω(m) after the timing synchronization loop is stabilized. When there is a timing error in the loop, ω(m) changes with the timing error value, and its calculation formula is:

[0118] ω(m+1)=ω(m)+C 1 (e(k)-e(k-1))+C 2 e(k-1) (1-31)

[0119] The time domain relationship between the interpolation controller register value and related parameters is as follows: Figure 7 shown.

[0120] Figure 7 Medium, m k T s With kT I They are the adjacent sampling time and interpolation time of the interpolation point, and the difference between them is μ k T s The register value η(m overflows at the moment when the interpolation controller outputs the interpolation enable signal. Figure 7 It can be concluded that

[0121]

[0122] Combining equation (1-29) with equation (1-32), we can solve the fractional interval

[0123]

[0124] The previous analysis shows that the interpolation filter based on the Farrow structure requires four sampling points per symbol, that is, the symbol period T = 4T s , the Gardner error detector requires two interpolation points for each symbol, that is, T = 2T i Therefore, ω(m)≈0.5, and μ k =2η(m k ), m k = int[kT i / T s ]=2k. Finally, the interpolation controller is based on m k and μ kThe interpolation point can be determined to stabilize the loop.

[0125] In order to verify the performance of the multi-level Gardner timing synchronization algorithm of the present invention, the timing synchronization loop was implemented using Matlab. First, the multi-level Gardner timing synchronization algorithm was functionally tested. In order to facilitate the observation of the correctness of its function, the input signal was set to a 0.5MHz sine signal to replace the time domain waveform of the signal, which is equivalent to a 1, -1 alternating sequence with a symbol rate of 1MHz. The sampling frequency was set to 4MHz, the frequency deviation was 1kHz, and the phase deviation was π / 3. The test results are as follows: Figure 8 .

[0126] from Figure 8 (a) It can be seen that the sampled signal contains four sampling points per symbol, and the interpolation filter outputs two points per symbol. After 20 sampling points, the loop reaches stability. The interpolation output points are also consistent with the theory, which are the best sampling point and the center point of the two best sampling points. Figure 8 (b) is the convergence curve of important parameters of the multi-level Gardner timing synchronization algorithm.

[0127] Since the sampling frequency has a frequency deviation of 1kHz, the fractional interval curve is a sawtooth waveform. The timing error quickly converges to 0, indicating that the timing deviation has been eliminated. The output ω(m) of the loop filter also converges to 0.5, which is consistent with the theoretical calculation. Based on the above analysis, it is verified that the timing synchronization algorithm invented in this paper is functionally correct.

[0128] The performance simulation of the multi-level Gardner timing synchronization algorithm is tested from the constellation diagram. From the constellation diagram, we can intuitively see the phase noise, amplitude noise, bit error, and whether the signal is synchronized. First, the constellation diagram test is performed on the four modulation signals of the meteorological satellite digital receiver standard. The test conditions are symbol rate of 1MHz, sampling frequency of 4MHz, normalized frequency deviation equal to 0.001, that is, 1kHz, phase deviation π / 2, SNR = 15dB, roll-off factor α = 0.35, and the test results are as follows Figures 9 to 12 .

[0129] From the simulation test results, it can be seen that the constellation diagram before timing synchronization adds clock frequency deviation and phase deviation, the symbols are aliased, the constellation diagram is messy, and the signal is not synchronized. After the timing synchronization algorithm, the symbols of the constellation diagram are regularly distributed.

Claims

1. A meteorological satellite digital receiver system, characterized in that: It includes an analog down-conversion module, an analog / digital converter (ADC) module, a digital down-conversion module, a matched filter module, timing synchronization, carrier synchronization, frame synchronization, equalization and automatic gain control (AGC), demapping, LDPC and BCH decoding, and a baseband data processing module. The radio frequency signal emitted by the satellite is received by the antenna and transmitted to the analog down-conversion module to be converted into an intermediate frequency analog signal, and then the intermediate frequency analog signal is sent to the ADC module to be converted into an intermediate frequency digital signal; the collected intermediate frequency digital signal is multiplied by the local oscillator signal to complete the digital down-conversion and output a zero intermediate frequency signal; the zero intermediate frequency signal is filtered by the matched filter module; the filtered signal is processed by timing synchronization, carrier synchronization and frame synchronization, and the data after carrier synchronization will be equalized and AGC processed. Finally, the decoded data is obtained by demapping, LDPC and BCH decoding in sequence, and is transmitted to the baseband data processing module to obtain a TS data stream, and the data after frame synchronization directly enters the baseband data processing module.

2. A meteorological satellite digital receiver system according to claim 1, characterized in that: The timing synchronization is used to ensure that the clocks of the receiving end and the transmitting end remain synchronized; the function of the carrier synchronization is to ensure that the local oscillator of the receiving end is synchronized with the carrier frequency and phase of the transmitting end.

3. A meteorological satellite digital receiver system according to claim 1 or 2, characterized in that: The timing synchronization adopts the multi-level Gardner timing synchronization algorithm, which includes an interpolation filter, a multi-level Gardner timing error detector, a loop filter and an interpolation controller; the input data x(mT s ) enters the interpolation filter, and the interpolation filter outputs the sampled data y(kT i ), and also outputs QPSK, 8PSK, 16APSK, and 32APSK signals, which enter the loop filter after passing through the multi-level Gardner error detection algorithm, and then enter the difference controller after the output signal from the loop filter.

4. A meteorological satellite digital receiver system according to claim 3, characterized in that: The interpolation filter is used to estimate other sampling points between the sampling points of the discrete signal, so as to achieve the conversion of the data rate; the interpolation filter includes a digital / analog converter, a filter, an analog-to-digital converter, and the data x(mT s ) is successively passed through the digital / analog converter, filter, and analog-to-digital converter to obtain the interpolation output y(kT i ).

5. A meteorological satellite digital receiver system according to claim 4, characterized in that: The interpolation filter adopts a cubic interpolation algorithm based on a Farrow structure.

6. A meteorological satellite digital receiver system according to claim 3, characterized in that: The multi-level Gardner timing error detection algorithm includes the following formula: In the formula [y I (k)+y I (k-1)]·β is the sample point at the middle moment between two symbols, y I (k), y I (k-1) represents the kth and k-1th symbols respectively, β = g T (T / 2) is the impulse response of the pulse shaping filter at time T / 2; In the meteorological satellite digital receiver system, the shaping filter uses a root raised cosine filter (RRC), and its impulse response is: In the formula, α is the roll-off factor of the filter. Substituting t=T / 2 into formula (1-15), β can be derived as: The specific value of β depends on α. ​​The meteorological satellite digital receiver standard stipulates that the value of α is 0.35, 0.25 and 0.

2. Substituting it into formula (1-16), the corresponding β values ​​are 0.619, 0.627 and 0.63 respectively.

7. A meteorological satellite digital receiver system according to claim 3, characterized in that: The loop filter is used to filter and process the feedback signal of the digital control loop; the loop filter adopts an ideal integral filter.

8. A meteorological satellite digital receiver system according to claim 3, characterized in that: The function of the interpolation controller is to generate an overflow clock signal, thereby controlling the interpolation filter to perform interpolation at an optimal sampling time.

Images