Communication satellite motion state estimation method based on Doppler frequency detection

Abstract

The invention belongs to the field of satellite communication, and relates to a communication satellite motion state estimation method based on Doppler frequency detection. The invention provides a method for estimating the motion state of a satellite through a satellite signal under a non-cooperative condition. According to the method, the motion state of the satellite is estimated by observing the Doppler effect generated when the satellite transmits the signal and the receiving end has relative motion. In order to accurately measure the Doppler frequency of the received satellite signal, the internal structure of a certain satellite signal needs to be mastered, and the invention provides a method for searching the internal structure of the signal based on a cross-correlation function of a communication signal. The method has the advantages that under the non-cooperative condition, blind analysis is carried out on the structure of the communication satellite signal through correlation calculation according to knowledge of the modern communication technology. And obtaining a specific repetitive sequence capable of calculating the Doppler frequency caused by the satellite motion according to an analysis result.

Description

Technical Field

[0001] The present invention belongs to the field of satellite communication and relates to a method for estimating the motion state of a communication satellite based on Doppler frequency detection. Background Art

[0002] Communication satellites play a crucial role in modern society and can achieve data transmission and communication globally. Understanding the motion state information of each communication satellite is helpful for aspects such as communication quality and information security. The difficulty in this regard lies in how to measure the motion state of a communication satellite under non-cooperative conditions.

[0003] Obtaining the signals transmitted by a satellite through radio monitoring is a reliable source of information for detecting the motion state of the satellite under non-cooperative conditions. This technical direction has the following advantages: strong real-time performance, currently calculations can achieve real-time monitoring of satellite signals, continuously receive the signals transmitted by the satellite, and timely process and analyze the data; not affected by weather, whether it is sunny or rainy, in most cases satellite signals can be received by the ground; wide coverage range, as long as the ground monitoring stations are reasonably set, global satellites can be monitored.

[0004] An important basis for calculating the motion state of a communication satellite based on the received signal is the Doppler effect. The Doppler effect refers to the change in the frequency of the wave received by the receiver when there is relative motion between the wave source and the receiver. If there is relative motion between the satellite and the ground receiving station, the frequency of the radio signal received by the receiving station will change, and this frequency change amount is the Doppler frequency shift. That is, the motion state of the satellite can be inferred based on the frequency change of the received signal.

[0005] Since electromagnetic waves propagate at the speed of light and the motion speed of the satellite is much less than the speed of light, the generated Doppler frequency shift is very small and the change in frequency cannot be directly observed by the short-time Fourier transform. In this task scenario, a method with higher precision is required to calculate the Doppler frequency, so measuring the phase change information of the signal is the main way to estimate the Doppler frequency.

[0006] Most of the existing methods for estimating the Doppler frequency based on phase change are under cooperative conditions and require obtaining the information of the electromagnetic waves transmitted by the satellite in advance to estimate its motion state, which does not meet the requirements of the preset task scenario. Summary of the Invention

[0007] The object of the present invention is to provide a method for estimating the motion state of a communication satellite based on Doppler frequency detection in view of the above problems. This method analyzes the structure of the received signal, extracts specific repeated sequences therein, and estimates the motion state of the communication satellite by calculating and analyzing the phase difference between the IQ two-channel signals of adjacent repeated sequences caused by the Doppler effect. Without prior knowledge of the target signal, this method can accurately estimate the Doppler change information from complex signals, thereby detecting the motion state of an object in a complex electromagnetic environment.

[0008] The technical solution of the present invention is as follows:

[0009] An estimation method for the motion state of a communication satellite based on Doppler frequency detection, defining the received satellite signal as a complex signal S c (t) with the sampling rate of F S , and the method includes:

[0010] S1. Blind extraction of the internal structure of the satellite signal:

[0011] The satellite signal S c (t) generally consists of a modulation signal, a synchronization signal, and a pilot signal. In the time domain, the signals of these three parts are independent of each other. Since the synchronization signal and the pilot signal are generally short-interval repeated sequences, they can be used for Doppler frequency detection.

[0012] First, normalize the received signal S c (t):

[0013]

[0014] And filter out the noise with a digital filter.

[0015] Without knowing the structure of the signal S c (t), in order to find the specific repeated sequence in the signal, calculate the autocorrelation of the entire signal S c (t):

[0016]

[0017] where T represents the offset length of two signals in the time domain, and the value range is ±t 0 , t 0 is the length of the signal S c (t), and the operation symbol |*| represents taking the modulus of a complex number.

[0018] For the obtained cross-correlation result Corr all(T), the highest peak will appear at T = 0. And some other secondary peaks are signs of the repetitive structure of the signal. Assume that a set of correlation peaks of the cross-correlation result appear at T = ±t 1 positions, which indicates that the signal may have a highly correlated structure at an interval of t 1 Record the positions where several of the highest secondary peaks other than the main peak appear as the intervals for subsequent searching of specific repetitive sequences. The cross-correlation result Corr all (T) finds the candidate intervals t 1 , t 2 , t 3 .....

[0019] S2. Determination of the specific repetitive sequence of the Doppler effect

[0020] Based on these intervals, find the positions and lengths where the repetitive sequence appears. Taking the interval t 1 as an example.

[0021] Assume that when t ∈ (t a , t b ), the signal S c (t) has continuous valid signals. Slice the signal at an interval of t 1 to obtain some adjacent sub-signals with a specific length, S c (t a : t a + t 1 ), S c (t a + t 1 : t a + 2 * t 1 )...... S c (t a + n * t 1 : t a +(n + 1) * t 1 ), where (n + 1) * t 1 ≤ t b .

[0022] Use a sliding window to calculate the cross-correlation values of adjacent sub-signals within the window:

[0023]

[0024] where window refers to the window length, generally taking 80 sampling points. k is less than n, and by adjusting the size of k, all adjacent sequences are traversed. T is less than t 1 - window, controlling the sliding of the window. The operation symbol |*| represents taking the modulus of a complex number.

[0025] The output result indicates whether there is a correlation between two sub-signal sequences adjacent to the window at the corresponding T position. If there is a repeated sequence of a certain length in two adjacent sub-signal sequences, it can be manifested as a continuous high-correlation value ladder in the function. The position where the ladder appears and the corresponding length represent the position and the corresponding length where the repeated sequence appears.

[0026] Process other intervals in accordance with the processing method for the interval t 1 ;

[0027] By comprehensively comparing the interval time when the repeated sequence appears and the length of the repeated sequence itself, select a suitable repeated sequence to calculate the Doppler frequency. The formula for the phase difference is as follows:

[0028]

[0029] where f t represents the Doppler frequency, and t represents the interval time. The phase difference between adjacent sequences will be ambiguous after exceeding π. Therefore, the interval when the repeated sequence appears should be short, so as to make the resolution of calculating the change in Doppler frequency larger. And the length of the repeated sequence should be appropriately long, and the phase difference is calculated through more samples, so as to make the error of the Doppler frequency smaller.

[0030] S3. Calculate the Doppler frequency of adjacent repeated sequences according to the phase difference

[0031] According to the comprehensive evaluation in the previous step, select a suitable repeated sequence. According to the result of Corr t1,k (T), extract a repeated sequence S c (t) from the overall signal S csub (t) ∈ S c (t). Use this repeated sequence S csub (t) to perform a cross-correlation calculation with the overall signal S c (t):

[0032]

[0033] where t l represents the length of the repeated sequence S csub (t), and T represents the position where the correlation function slides.

[0034] When changing T such that the repeated sequence S csub (t) completely overlaps with other repeated sequences in S c (t), the function Corr sub (T) will have a peak, representing the position where the repeated sequence appears, and record these positions.

[0035] In theory, all the repeated sequences at the transmitting end should be exactly the same. However, due to the influence of the Doppler effect, the frequency of the repeated sequences is changed. By calculating the phase difference between the I and Q signals of adjacent repeated sequences, the Doppler frequency with a relatively small magnitude can be calculated. Let two adjacent repeated sequences be S csub,1 (t) and S csub,2 (t). The specific formula for calculating the Doppler frequency f t is as follows:

[0036]

[0037] where t s represents the total duration of the repeated sequence, and the operator θ(*) represents the phase angle of the complex number, which calculates the phase angle of the IQ signal at time t in the formula.

[0038] S4. Fitting the satellite motion state according to the Doppler frequency

[0039] After calculating the Doppler frequency f t , the motion state of the communication satellite can be estimated according to the formula of the Doppler effect. Estimate the velocity v of the target object at the current moment:

[0040] v = λf t / 2

[0041] where v represents the velocity of the target, λ is the wavelength of the electromagnetic wave emitted by the target, and f t represents the measured Doppler frequency.

[0042] Moreover, the received signal will have repeated sequences at relatively short intervals (generally on the order of milliseconds). By estimating the target velocity at the current moment, the target velocity information at the current moment with a short interval can be obtained.

[0043] Fitting the target velocity information obtained at all moments to obtain the time-velocity curve of the target, so as to estimate the motion state of the target object. When the time-velocity curve is a flat straight line, it indicates that there is no obvious change in the radial relative velocity of the target; when the time-velocity curve has obvious fluctuations, the fluctuation process represents the change process of the radial velocity.

[0044] The beneficial effects of the present invention are as follows: In a non-cooperative situation, through correlation calculation and based on the knowledge of modern communication technology, the present invention performs blind analysis on the structure of the communication satellite signal. According to the analysis results, specific repeated sequences capable of calculating the Doppler frequency caused by the satellite motion are obtained. Finally, the motion state of the communication satellite is estimated based on the received communication satellite signal.

[0045] The object of the present invention is to judge the motion state of a communication satellite based on the received communication satellite signal without other prior conditions. The motion state of the satellite can be estimated by the Doppler frequency change of the signal caused by the motion of the communication satellite. Description of the Drawings

[0046] Figure 1 Flow chart of the technical solution;

[0047] Figure 2 Example of the received satellite signal, including the time-domain diagrams of the I and Q signals and the STFT diagrams;

[0048] Figure 3 Autocorrelation function diagram of the satellite signal;

[0049] Figure 4 Sliding window cross-correlation diagram of adjacent sequences of a specific length;

[0050] Figure 5 Partial intercepted diagram of the cross-correlation function between the repeated sequence and the satellite signal;

[0051] Figure 6 Doppler frequency diagram of the received real signal a;

[0052] Figure 7 Doppler frequency diagram of the received real signal b;

[0053] Figure 8 Doppler frequency diagram of the simulated signal c;

[0054] Figure 9 Doppler frequency diagram of the simulated signal d. Detailed Implementation Manner

[0055] The technical solution of the present invention will be further described below in conjunction with the drawings and simulations.

[0056] In this example, the calculation and fitting process of the proposed method will be introduced, taking a section of the received real satellite signal as an example. The satellite signal is 1 s long and the sampling frequency is 56 mHz. Figure 2 are the time-domain diagrams of the partial I and Q signals and the STFT diagrams of the signal.

[0057] In order to detect whether there are characteristic repeated sequences in the signal that can be used to detect the phase difference, the autocorrelation of the entire signal is calculated, and the result is as shown in Figure 3 . It can be seen from the figure that every 28,000 sampling points, the function will have a high correlation peak, indicating that the received signal will have a specific repeated sequence every 28,000 sampling points.

[0058] According to Figure 3As a result, two adjacent sequences of 28,000 sampling points in the signal were taken, and a sliding window of 80 sampling points was used to calculate the cross-correlation of the two sequences, obtaining the result as shown in Figure 4 . It can be seen that Figure 4 there are obvious steps from the 18,000th to the 22,000th sampling points, indicating that at these corresponding sampling point positions, two adjacent signals are highly correlated, and it can be considered that the signals appearing at the corresponding positions are specific repeated sequences.

[0059] According to Figure 4 the result, the repeated sequence was extracted, and the cross-correlation between the repeated sequence and the entire signal was calculated. Figure 5 The partial cross-correlation results are shown. A high correlation peak appears every 28,000 sampling points in the cross-correlation results, and the positions where these correlation peaks appear are the exact positions where the repeated sequences appear.

[0060] According to Figure 4 the length of the repeated sequence was obtained, and Figure 5 the positions where the repeated sequences appear. Each repeated sequence was extracted and the phase difference between adjacent sequences was calculated, and the Doppler frequency at the corresponding time position was calculated, thus obtaining the Doppler curve as shown in Figure 6 .

[0061] Figure 6 is the Doppler frequency diagram within 1 s of a received real satellite signal. The repeated sequence appears every 28,000 sampling points in the signal, so the Doppler frequency can be estimated every t = 28,000 / 56,000,000 = 0.5 ms. The Doppler frequency can reflect the radial motion speed between the satellite and the receiving end, Figure 6 and the fitting results in

[0062] Figure 7 show that the motion state of the satellite is changing.

[0063] Figure 8 , Figure 9 in which the red curve represents the Doppler frequency to be simulated. A broadband signal was applied with the corresponding Doppler frequency as the simulation input, and this signal was subjected to Doppler frequency fitting according to the process of the present invention, obtaining the blue broken line in the figure. It can be seen that the present invention can effectively measure the Doppler curve, thereby obtaining the motion state of the object.

Claims

1. A method for estimating the motion state of a communication satellite based on Doppler frequency detection, defining the received satellite signal as a complex signal S containing two paths: IQ c (t), the sampling rate is F S , characterized in that, The following steps are involved: S1, for the received signal S c (t) After normalization and filtering out noise with a digital filter, the signal S is calculated. c Autocorrelation of (t): Where T represents the offset length of the two signals in the time domain, and the value range is ±t0, t0 is the signal S c (t) is the length of the complex number. The operator |*| represents the modulus of the complex number. For the cross-correlation result Corr all (T), the highest main peak appears at T = 0, and a group of correlation peaks of the cross-correlation result appear at T = ± t1, indicating that the signal has a highly correlated structure in the interval t1. The positions of several highest secondary peaks except the main peak are recorded and defined as the candidate intervals t1, t2, t3... where the repeated sequence appears; S2. Find the position and length of the repeated sequence according to the selected interval where the repeated sequence appears. The processing method for the interval t1 is: Define t∈(t a ,t b ), the signal S c (t) There is a continuous valid signal. The signal is sliced ​​according to the interval t1 to obtain adjacent sub-signals, which are expressed as S c (t a :t a +t1),S c (t a +t1:t a +2*t1)......S c (t a +n*t1:t a +(n+1)*t1), where (n+1)*t1≤t b ; Use the sliding window to calculate the cross-correlation value of adjacent sub-signals within the window: Where window refers to the window length, k is less than n, and all adjacent sequences are traversed by adjusting the size of k. T is less than t1-window to control the sliding of the window. The operator |*| represents the modulus of a complex number. The output result indicates whether the two adjacent sub-signal sequences in the window at the corresponding T position are correlated. If there are repeated sequences in the two adjacent sub-signal sequences, it can be The function is expressed as a continuous high correlation value ladder, and the position and corresponding length of the ladder represent the position and corresponding length of the repeated sequence; The other intervals are processed according to the processing method of interval t1 to obtain the positions and corresponding lengths of all repeated sequences; By comparing the interval time of the repetitive sequence and the length of the repetitive sequence itself, the repetitive sequence that meets the set requirements is selected to calculate the Doppler frequency. The formula for the phase difference is as follows: where f t represents the Doppler frequency, t represents the interval time, and the phase difference between adjacent sequences When the number exceeds π, there will be ambiguity. Set the requirements for the required repetitive sequence, and then select the repetitive sequence S that meets the requirements. csub (t); S3, using repeated sequence S csub (t) and the overall signal S c (t) Calculate the cross-correlation: where t l Represents a repeating sequence S csub The length of (t), T represents the position where the correlation function slides; When changing T so that the repeating sequence S csub (t) and S c When the other repeated sequences in (t) completely overlap, the function Corr sub (T) Peaks will appear, representing the locations where the repeated sequences appear, and these locations are recorded; Define two adjacent repeating sequences as S csub,1 (t) and S csub,2 (t), specific Doppler frequency f t The calculation formula is as follows: where t s represents the total duration of the repetitive sequence, and the operator θ(*) represents the phase angle of the complex number. The phase angle of the IQ signal at time t is calculated in the formula; S4. Get the Doppler frequency f t Then, the motion state of the communication satellite is estimated according to the formula of the Doppler effect, and the velocity v of the target object at the current moment is estimated: v=λf t / 2 Where v represents the speed of the target, λ is the wavelength of the electromagnetic wave emitted by the target, and f t represents the measured Doppler frequency; The received signal will have a repetitive sequence in a short interval time, and the target speed information of the current moment can be obtained by estimating the target speed at the current moment in a short interval, and the short interval time refers to the time in the order of milliseconds; The target speed information obtained at all times is fitted to obtain the target's time-speed curve, thereby estimating the motion state of the target object.

Images