Low-orbit opportunity signal Doppler estimation method for multi-beam dynamic switching

Abstract

The invention discloses a low-orbit opportunity signal Doppler estimation method for multi-beam dynamic switching, and belongs to the technical field of satellite positioning and navigation. The method comprises the following steps: firstly, carrying out blocking and band-pass filtering preprocessing on an iridium time frame signal according to a beam polling period, then, carrying out Doppler frequency coarse estimation by utilizing segmented FFT (Fast Fourier Transform), and carrying out accurate positioning on an initial position of a pilot signal in combination with locally generated priori independent word information. Secondly, extracting two-dimensional parameters such as frequency deviation factors and signal-to-noise ratio characteristics for the pilot signals subjected to accurate stripping, performing dynamic decision making by using a pre-trained KNN model, and performing Doppler frequency fine estimation by selecting a Rife or MLE algorithm according to the dynamic decision making. According to the invention, through an adaptive algorithm matching mechanism, the problem of insufficient precision of a traditional method in complex scenes of large signal-to-noise ratio fluctuation and large frequency change is solved, the robustness of Doppler estimation is significantly improved, and the method is suitable for non-cooperative signal processing of a low-orbit satellite navigation system.

Description

Technical Field

[0001] The present invention belongs to the technical field of satellite positioning and navigation, and in particular relates to a Doppler estimation method for low-orbit opportunity signals facing multi-beam dynamic switching. Background Art

[0002] The Global Navigation Satellite System (GNSS) is currently the most widely used navigation and positioning technology. However, its low signal power and susceptibility to interference and obstruction can lead to service degradation or even failure in complex environments. Low-Earth Orbit (LEO) satellites, such as the Iridium system, are an important backup or augmentation tool in GNSS-denied environments due to their high landing power (approximately 30dB stronger than GNSS), wide coverage, and rapidly changing geometry. However, LEO satellite signals exhibit significant Doppler shift (up to ±40kHz), and dynamic beam switching causes significant fluctuations in signal-to-noise ratio, posing challenges for signal processing.

[0003] Existing Doppler estimation methods for LEO non-cooperative signals include phase-time algorithms, QSA-IDE algorithms, and implicit pilot extraction algorithms. These algorithms have the following shortcomings: First, they rely on prior information, while LEO signals of opportunity often lack stable pilots or frame structures; second, their high algorithm complexity makes it difficult to achieve efficient processing within millisecond-scale signal frames; and third, their poor adaptability to signal-to-noise ratio fluctuations and large dynamic Doppler ranges leads to unstable estimation accuracy. These shortcomings limit the practical application of LEO signals of opportunity in navigation and positioning.

[0004] Therefore, there is an urgent need for a robust estimation method that can adapt to the large dynamic Doppler and signal-to-noise ratio fluctuations of low-orbit signals, and achieve high-precision and high-efficiency Doppler frequency estimation in the absence of prior information, so as to improve the availability of low-orbit opportunity signals in complex environments. Summary of the Invention

[0005] To address the aforementioned technical issues, the present invention provides a Doppler estimation method for low-orbit signals of opportunity (LEOs) with dynamic multi-beam switching. Based on the Iridium satellite time-frame pilot signal, a global segmented FFT is used to perform a coarse Doppler frequency estimate. Frame header synchronization is then performed using prior information from the Iridium satellite frame to identify and extract characteristic parameters such as the frequency offset factor and signal-to-noise ratio (SNR). This characteristic parameter pair is then input into a dynamic decision model using a pre-trained KNN (K Near Neighbor) algorithm in a two-dimensional feature parameter space (frequency offset factor, instantaneous SNR). A "parameter space-refinement algorithm" matching mapping mechanism is established, dynamically coupling the algorithm architecture with signal characteristics. This method enables adaptive algorithm selection based on the complex characteristics of complex scenarios, optimizing Doppler estimation accuracy and robustness.

[0006] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0007] In a first aspect, the present invention provides a Doppler estimation method for low-orbit signal of opportunity (LOO) for dynamic multi-beam switching, comprising:

[0008] Step 1: Divide the received LEO opportunity signal into blocks according to a preset period, and pre-process each block signal to obtain a target time frame signal, wherein the target time frame signal includes at least a pilot and a unique word of the LEO opportunity signal; wherein the preset period matches the beam polling period of the LEO opportunity signal;

[0009] Step 2: Segment the target time frame signal, and perform fast Fourier transform (FFT) on each segment signal of the target time frame signal to obtain the frequency domain peak value of each segment signal, determine the segment signal whose frequency domain peak value meets the preset peak value condition as the target segment signal, and determine the Doppler frequency rough estimation value of the pilot signal based on the frequency domain peak value in the target segment signal; wherein, the preset peak value condition includes that the frequency domain peak value of the current segment signal is greater than a preset peak value threshold, and after sorting the frequency domain peak values ​​of all segment signals from large to small, the frequency domain peak value of the current segment signal is located in the first N digits, where N is a positive integer;

[0010] Step 3: Generate a local independent word signal using prior information, determine the peak of the cross-correlation function between the local independent word signal and the panned pilot coarse position signal as the pilot start position, and the panned pilot coarse position signal is a signal whose distance from the pilot coarse position is less than a preset distance threshold; wherein the prior information includes at least prior pilot information and prior unique word information;

[0011] Step 4: Determine a pilot signal based on the pilot starting position, perform FFT processing on the pilot signal again to obtain a pilot signal spectrum, and determine a frequency offset factor estimate and an instantaneous signal-to-noise ratio estimate based on the pilot signal spectrum; wherein the number of points in the FFT processing again is the same as the number of time-domain sampling points of the pilot signal;

[0012] Step 5: Inputting the frequency offset factor estimate and the instantaneous signal-to-noise ratio estimate into a pre-trained KNN decision model to obtain a first algorithm weight and a second algorithm weight; wherein the first algorithm is a frequency estimation Rife algorithm based on discrete-time signal phase difference, and the second algorithm is a maximum likelihood estimation MLE algorithm;

[0013] Step 6: Use the target algorithm to determine the Doppler frequency compensation value of the pilot, and determine the precise value of the Doppler frequency of the pilot based on the rough estimate of the Doppler frequency of the pilot and the Doppler frequency compensation value of the pilot; wherein, when the weight of the first algorithm is greater than the weight of the second algorithm, the target algorithm is the first algorithm; when the weight of the second algorithm is greater than the weight of the first algorithm, the target algorithm is the second algorithm.

[0014] The beneficial effects of the present invention are:

[0015] High Precision and Robustness: This method extracts signal characteristic parameters (frequency offset factor and signal-to-noise ratio) and leverages a pre-trained KNN model to dynamically select the optimal estimation algorithm (Rife or MLE), achieving complementary algorithmic advantages. Experiments show that this method significantly reduces root mean square error (RMS) and improves estimation accuracy by over 30% under conditions of signal-to-noise ratio fluctuations (-15dB to 5dB) and large Doppler dynamics (±40kHz).

[0016] Adaptive scene matching capability: By building a "feature parameter-algorithm" mapping mechanism, this invention can automatically select the optimal processing path based on the real-time signal characteristics. For example, when the signal-to-noise ratio is high, the computationally less complex Rife algorithm is prioritized, while when the signal-to-noise ratio is low, the more noise-resistant MLE algorithm is switched, thus achieving a balance between accuracy and efficiency.

[0017] Low complexity and high real-time performance: Compared to algorithms such as the phase-time method that require a full-band search, this method significantly reduces computational complexity by performing a step-by-step process of coarse and fine estimation. Furthermore, Doppler estimation can be performed using only pilot signals (without demodulating data frames), avoiding the complex demodulation process required by implicit pilot methods. This improves computational efficiency by over 50%, making it suitable for scenarios requiring high real-time performance.

[0018] Wide applicability: The present invention does not rely on prior information about a specific satellite system and can be directly applied to low-orbit opportunity signals such as Iridium. It is also compatible with non-cooperative signal processing of other similar systems, providing reliable technical support for navigation and positioning in GNSS-denied environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 The present invention provides a flowchart of a method for estimating the Doppler of a low-orbit opportunity signal for dynamic multi-beam switching.

[0020] Figure 2 This is a schematic diagram of the time frame structure of the Iridium signal.

[0021] Figure 3 The figure is a flow chart of a method for determining a rough Doppler frequency estimation value of a pilot signal provided by an embodiment of the present invention.

[0022] Figure 4It is a flowchart of a method for determining a pilot starting position provided by an embodiment of the present invention.

[0023] Figure 5 This is a schematic diagram of the pilot signal determined using Iridium as an example.

[0024] Figure 6 4 is a flow chart of a method for training a KNN decision model provided by an embodiment of the present invention.

[0025] Figure 7 This is a schematic diagram of a set of label points in a pre-trained KNN decision model provided by an embodiment of the present invention.

[0026] Figure 8 It is a flowchart of a method for determining the weight of a first algorithm and the weight of a second algorithm using a pre-trained KNN decision model provided by an embodiment of the present invention.

[0027] Figure 9 This is a flow chart of another method for Doppler estimation of low-orbit opportunity signals for dynamic multi-beam switching provided by the present invention. DETAILED DESCRIPTION

[0028] The present invention will be further described below with reference to the accompanying drawings and examples.

[0029] Figure 1 This is a flow chart of a method for estimating the Doppler of a low-orbit opportunity signal with dynamic multi-beam switching provided by the present invention. Figure 1 As shown, the method includes the following steps:

[0030] In step 1, the received low-orbit opportunity signal is divided into blocks according to a preset period, and each block signal is preprocessed to obtain the target time frame signal.

[0031] The target time frame signal includes at least a pilot and a unique word of the low-orbit opportunity signal, and the preset period matches the beam polling period of the low-orbit opportunity signal.

[0032] In step 2, the target time frame signal is segmented and the fast Fourier transform (FFT) is performed on each segmented signal of the target time frame signal to obtain the frequency domain peak of each segmented signal. The segmented signal whose frequency domain peak meets the preset peak condition is determined to be the target segmented signal, and the rough estimate of the Doppler frequency of the pilot is determined based on the frequency domain peak in the target segmented signal.

[0033] Among them, the preset peak condition includes that the frequency domain peak of the current segment signal is greater than the preset peak threshold, and after sorting the frequency domain peaks of all segment signals from large to small, the frequency domain peak of the current segment signal is in the first N positions, where N is a positive integer.

[0034] In step 3, a local independent word signal is generated using prior information, and the peak of the cross-correlation function between the local independent word signal and the pan-pilot coarse position signal is determined as the pilot start position.

[0035] The pan-pilot coarse position signal is a signal whose distance from the pilot coarse position is less than a preset distance threshold; the prior information at least includes prior pilot information and prior unique word information.

[0036] In step 4, the pilot signal is determined based on the pilot starting position, and the pilot signal is subjected to FFT processing again to obtain the pilot signal spectrum. The frequency offset factor estimation value and the instantaneous signal-to-noise ratio estimation value are determined based on the pilot signal spectrum.

[0037] The number of points in the re-performed FFT processing is the same as the number of time-domain sampling points of the pilot signal.

[0038] In step 5, the frequency offset factor estimation value and the instantaneous signal-to-noise ratio estimation value are input into a pre-trained KNN decision model to obtain a first algorithm weight and a second algorithm weight.

[0039] Among them, the first algorithm is the frequency estimation Rife algorithm based on the phase difference of discrete time signals, and the second algorithm is the maximum likelihood estimation MLE algorithm.

[0040] In step 6, a target algorithm is used to determine a Doppler frequency compensation value of the pilot, and an accurate Doppler frequency value of the pilot is determined based on a rough estimation value of the Doppler frequency of the pilot and the Doppler frequency compensation value of the pilot.

[0041] Among them, when the weight of the first algorithm is greater than the weight of the second algorithm, the target algorithm is the first algorithm; when the weight of the second algorithm is greater than the weight of the first algorithm, the target algorithm is the second algorithm.

[0042] In some embodiments of the present invention, the method can be executed by a server or a terminal device with certain processing capabilities. In some implementations, the method can be used to accurately estimate the Doppler spectrum of low-orbit satellite opportunity signals, and then use the estimation results to assist in positioning and navigation.

[0043] In some embodiments of the present invention, a low-orbit opportunity signal may be received first and then divided into blocks according to a preset period. The low-orbit opportunity signal may be a low-orbit satellite opportunity signal. In one example, the low-orbit satellite opportunity signal may be an Iridium satellite signal.

[0044] The preset period matches the beam polling period of the low-orbit signal of opportunity, ensuring that each segmented signal includes a pilot segment. Furthermore, the pilot signal, an unmodulated continuous wave signal, can be received, processed, and accurately estimated without prior information. Doppler frequency estimation can be performed by obtaining the pilot signal from each segmented signal.

[0045] Taking the Iridium signal as an example, since the Iridium system polls 48 beams with a period of 4.32 seconds (s) when broadcasting signals, the acquired Iridium downlink signal can be divided into blocks according to the signal broadcast period, with each 4.32s signal as a block, thereby ensuring that the Iridium time frame signal exists in each signal block.

[0046] In some embodiments of the present invention, each block signal after block completion may be preprocessed to obtain a target time frame signal therein, where the target time frame signal includes at least a pilot and a unique word of the low-orbit opportunity signal.

[0047] Still taking the Iridium signal as an example, Figure 2 This is a schematic diagram of the time frame structure of the Iridium signal. Figure 2 As shown in the figure, the Iridium signal time division multiple access (TDMA) frame length is 90 milliseconds (ms), of which the simplex channel is allocated a time slot of 20.32ms, the duplex channel occupies 4 uplink channels of 8.28ms and 4 downlink channels of 8.28ms, and the frame header is allocated a 1ms guard interval. The simplex channel and the duplex uplink channel are allocated a 1.24ms guard interval. The duplex uplink channel is allocated a 0.22ms guard interval between each time slot. The duplex downlink channel is allocated a 0.1ms guard interval between each time slot. The duplex uplink channel and the duplex downlink channel are allocated a 0.24ms guard interval.

[0048] A simplex channel can contain a pilot, a unique word, and data. The pilot occupies a 2.56ms time slot, the unique word consists of 12 characters, each of which occupies a 0.04ms time slot, for a total of 0.48ms, and the data occupies a 17.28ms time slot. The pilot is unmodulated, the unique word is modulated using binary phase shift keying (BPSK), and the data is modulated using quadrature phase shift keying (QPSK).

[0049] The simplex channel time frame may be used as the target time frame signal of the Iridium satellite signal, so that the target time frame signal includes the pilot and unique word of the Iridium satellite signal.

[0050] In certain embodiments of the present invention, the target time frame signal can be segmented and a Fast Fourier Transform (FFT) performed on each segment of the target time frame signal to obtain the frequency domain peak value of each segment. In one example, the target time frame signal of each signal block can be further divided into a number of segments of equal length. Each segment is then subjected to an FFT to obtain the spectrum of each segment, which at least includes the frequency domain peak value.

[0051] A segmented signal whose frequency domain peak value satisfies a preset peak condition may be determined as a target segmented signal, and a rough Doppler frequency estimate of the pilot signal may be determined based on the frequency domain peak value in the target segmented signal. The preset peak condition may include: the frequency domain peak value of the segmented signal being greater than a preset peak threshold, and the frequency domain peak value of the segmented signal being in the first N positions after sorting the frequency domain peak values ​​of all segmented signals from largest to smallest, where N is a positive integer.

[0052] In one example, the first N frequency domain peaks greater than a preset peak threshold among the frequency domain peaks of each segmented signal can be determined as frequency domain peaks that meet the preset peak condition. If the number of frequency domain peaks greater than the preset peak threshold is less than N, then all frequency domain peaks greater than the preset peak threshold are determined to be frequency domain peaks that meet the preset peak condition. The specific values ​​of the preset peak threshold and N can be set according to actual needs. For example, the preset peak threshold can be set to -15dB, and N can be set to 9.

[0053] In some embodiments of the present invention, a priori information may be obtained, including at least a priori pilot information and a priori unique word information. Furthermore, a local independent word signal may be generated using the a priori information, and the peak of the cross-correlation function between the local independent word signal and the rich pilot coarse position signal may be determined as the pilot starting position. The rich pilot coarse position signal is a signal whose distance from the pilot coarse position is less than a preset distance threshold.

[0054] In some embodiments of the present invention, a pilot signal can be determined based on the pilot starting position, and then the pilot signal can be subjected to a further FFT process to obtain the pilot signal spectrum. That is, after determining the pilot starting position, the pilot signal can be stripped from the pilot starting position based on the time slot length of the pilot signal in the LEO signal of opportunity. The number of points in the further FFT process is the same as the number of time-domain sampling points of the pilot signal.

[0055] Then, the frequency offset factor estimation value and the instantaneous signal-to-noise ratio estimation value can be determined based on the pilot signal spectrum. The specific implementation method of determining the frequency offset factor estimation value and the instantaneous signal-to-noise ratio estimation value is described in detail below and will not be repeated here.

[0056] In some embodiments of the present invention, a KNN decision model may be pre-trained to determine an algorithm for refining pilot estimation based on a frequency offset factor and an instantaneous signal-to-noise ratio. The calculated frequency offset factor estimate and instantaneous signal-to-noise ratio estimate may be input into the pre-trained KNN decision model to obtain a first algorithm weight and a second algorithm weight, where the first algorithm may be a frequency estimation Rife algorithm based on a discrete-time signal phase difference, and the second algorithm may be a maximum likelihood estimation (MLE) algorithm.

[0057] The first algorithm weight and the second algorithm weight are compared. If the first algorithm weight is greater than the second algorithm weight, the first algorithm is determined to be the target algorithm. Conversely, if the second algorithm weight is greater than the first algorithm weight, the second algorithm is determined to be the target algorithm. The determined target algorithm is used to determine a Doppler frequency compensation value for the pilot. Finally, the precise Doppler frequency value of the pilot is determined based on the rough Doppler frequency estimate of the pilot and the Doppler frequency compensation value of the pilot.

[0058] The technical solution provided by the embodiment of the present invention solves the problem of insufficient accuracy of traditional methods in complex scenarios with large signal-to-noise ratio fluctuations and large frequency changes through an adaptive algorithm matching mechanism, significantly improves the robustness of Doppler estimation, and is suitable for non-cooperative signal processing in low-orbit satellite navigation systems.

[0059] In some embodiments of the present invention, when acquiring the target time frame signal, the received LEO opportunity signal can be divided into fixed-duration block signals according to the beam polling period of the LEO opportunity signal. Each block signal is then bandpass filtered to obtain the target time frame signal.

[0060] Taking the Iridium satellite signal as an example, the bandwidth allocated to Iridium is 1616.0 MHz to 1626.5 MHz. The bands 1616.0 MHz to 1626.0 MHz are duplex channels used as traffic channels, while the bands 1626.0 MHz to 1626.5 MHz are simplex channels used as signaling channels. Therefore, the lower cutoff frequency of the bandpass filter can be set to 1626.0 MHz or a value less than 1626.0 MHz, and the upper cutoff frequency to 1626.5 MHz or a value greater than 1626.5 MHz. This allows the simplex channel containing the pilot and unique word to be filtered out of the Iridium satellite signal as the target time frame signal.

[0061] Figure 3 FIG. 1 is a flow chart of a method for determining a rough estimate of the Doppler frequency of a pilot signal provided by an embodiment of the present invention. Figure 3 As shown, the method includes the following steps:

[0062] In step 201, each block signal is divided into a plurality of signal segments of equal length, and each signal segment is independently subjected to FFT transformation to obtain the frequency domain peak value of each segment signal.

[0063] In step 202, target segmented signals are determined, and the frequency domain peak value of each target segmented signal is obtained.

[0064] In step 203, a coarse Doppler frequency estimate of the pilot is determined based on the coarse position of the pilot.

[0065] In some embodiments of the present invention, each block signal can be first divided into multiple equal-length signal segments. Then, an FFT transform can be performed independently on each signal segment to obtain the frequency domain peak value of each segmented signal. Then, based on the frequency domain peak value of each segmented signal, a target segmented signal can be determined within the multiple equal-length signal segments, and the frequency domain peak value of each target segmented signal can be obtained. The specific implementation method for determining the target segmented signal is described in detail above and will not be repeated here.

[0066] Finally, you can use Determine a coarse estimate of the Doppler frequency of the pilot based on the coarse position of the pilot, where: is the rough estimate of the Doppler frequency of the pilot, is the spectral line number corresponding to the maximum peak value of each target segment signal, is the sampling rate for FFT transformation of each signal segment, is the number of points used to perform FFT transformation on each signal segment.

[0067] Figure 4 FIG. 1 is a flow chart of a method for determining a pilot starting position provided by an embodiment of the present invention. Figure 4 As shown, the method includes the following steps:

[0068] In step 401, a local independent word signal is generated using a priori information.

[0069] In step 402, a cross-correlation function is constructed between the local independent word signal and the pan-pilot coarse position signal, and the peak of the cross-correlation function is determined as the pilot starting position.

[0070] In some embodiments of the present invention, a local independent word signal may first be generated using a priori information ,in is a local independent word signal, is the a priori pilot amplitude, is the a priori unique word amplitude, is the imaginary number symbol, is the rough estimate of the Doppler frequency of the pilot, is the sampling rate for FFT transformation of each signal segment, n is the time domain sampling point index of the pilot signal, and each n value represents a sampling point. is the number of sampling points for each independent symbol of each LEO signal of opportunity, is the a priori unique character symbol information, The symbol for rounding up.

[0071] Then, construct the cross-correlation function between the local independent word signal and the pan-pilot coarse position signal, and determine the peak of the cross-correlation function as the pilot starting position. ,in is the pilot starting position, The function with the independent variable k is The maximum independent variable point set function of is the cross-correlation function, is the point index of the cross-correlation function, for The number of points, is the coarse position signal of the flood pilot.

[0072] As before, after the pilot starting position is determined, the pilot signal can be obtained by stripping the pilot starting position. Figure 5 This is a schematic diagram of the pilot signal determined using Iridium as an example. Figure 5 As shown, each 4.32-second signal block includes at least one 90-ms Iridium signal frame. For example, signal block m includes two 90-ms Iridium signal frames, and signal block m+1 includes one 90-ms Iridium signal frame. The frames are chronologically numbered as frame j, frame j+1, and frame j+2. After determining the pilot starting position, a pilot signal can be stripped 2.56 ms after each pilot starting position to obtain pilot j, pilot j+1, and pilot j+2, respectively.

[0073] Each determined pilot signal can be subjected to FFT processing again to obtain the spectrum of each pilot signal. Furthermore, the frequency offset factor estimation value and the instantaneous signal-to-noise ratio estimation value can be determined using the pilot signal spectrum.

[0074] In some implementations, the frequency offset factor estimate may be determined in the following manner: First, determine the peak value of the pilot signal spectrum. , then extract the spectrum peak Two adjacent spectral lines and . Finally, use the formula:

[0075]

[0076] ;

[0077] The frequency offset factor estimate is calculated, where is the estimated value of the frequency offset factor, is the frequency estimation direction.

[0078] In some other implementations, the instantaneous signal-to-noise ratio estimate may be determined by first determining the peak value of the pilot signal spectrum. , and then use the formula The instantaneous signal-to-noise ratio estimate is calculated, where is the estimated instantaneous signal-to-noise ratio, is the number of points for the FFT processing to be performed again, is the spectral line index of the pilot signal, is the first spectral lines.

[0079] Figure 6 FIG. 1 is a flow chart of a method for training a KNN decision model according to an embodiment of the present invention. Figure 6 As shown, the method includes the following steps:

[0080] In step 11, a Monte Carlo simulation is performed within a preset frequency offset factor and a preset instantaneous signal-to-noise ratio range based on real Iridium satellite signal statistics to obtain the root mean square error of the Doppler frequency estimation accuracy when the first algorithm and the second algorithm are used respectively for different combinations of frequency offset factors and instantaneous signal-to-noise ratios.

[0081] In step 12, the algorithm label of each frequency offset factor and instantaneous signal-to-noise ratio combination is determined based on the precision error, and the label point in the pre-trained KNN decision model is obtained.

[0082] Each tag point includes a pair of frequency offset factor and instantaneous signal-to-noise ratio combination, and an algorithm tag of the tag point, where the algorithm tag is the first algorithm or the second algorithm.

[0083] In some embodiments of the present invention, a Monte Carlo simulation can be performed within a preset frequency offset factor and instantaneous signal-to-noise ratio range based on real Iridium satellite signal statistics to obtain the root mean square error (RMS) of Doppler frequency estimation accuracy when using the first algorithm and the second algorithm for different combinations of frequency offset factor and instantaneous signal-to-noise ratio. Based on the accuracy error, an algorithm label is then determined for each frequency offset factor and instantaneous signal-to-noise ratio combination to obtain a label point in the pre-trained KNN decision model.

[0084] Figure 7 Schematic diagram of a set of label points in a pre-trained KNN decision model provided by an embodiment of the present invention. Figure 7 As shown, each label point includes a frequency offset factor parameter value and an instantaneous signal-to-noise ratio parameter value, and also includes an algorithm label, which can be a Rife algorithm or an MLE algorithm.

[0085] Figure 8 1 is a flow chart of a method for determining the weight of the first algorithm and the weight of the second algorithm using a pre-trained KNN decision model provided by an embodiment of the present invention. Figure 8 As shown, the method includes the following steps:

[0086] In step 501, the frequency offset factor estimate and the instantaneous signal-to-noise ratio estimate are input into a pre-trained decision model, and the Euclidean distance between the combination of the frequency offset factor estimate and the instantaneous signal-to-noise ratio estimate and each label point is calculated.

[0087] In step 502, K label points with the smallest Euclidean distance are selected as the decision sample space.

[0088] In step 503, a first algorithm weight and a second algorithm weight are calculated based on at least the decision sample space.

[0089] In some embodiments of the present invention, when determining the first algorithm weight and the second algorithm weight, the frequency offset factor estimate and the instantaneous signal-to-noise ratio estimate can be first input into a pre-trained decision model, and the Euclidean distance between the combination of the frequency offset factor estimate and the instantaneous signal-to-noise ratio estimate and each label point can be calculated.

[0090] Then select the K label points with the smallest Euclidean distance as the decision sample space ,in, is the decision sample space, K is a positive integer.

[0091] Next, you can use the formula The first algorithm weight is calculated, where is the first algorithm weight, Is a logical judgment symbol, if If established The value is 1, otherwise it is 0. Represents the decision sample space Medium sample The algorithm label in the corresponding label point is the first algorithm, A non-zero decimal number used to avoid zero denominators in fractions.

[0092] That is to say, when calculating the weight of the first algorithm, we can first select all the label points whose algorithm label is the first algorithm in the decision sample space, and then calculate the Euclidean distance and non-zero minimum value of each label point to the decision point. The reciprocal of the sum of the values ​​of the two labels is calculated, and finally the reciprocals of the two labels are added together to obtain the weight of the first algorithm. The decision point is determined by the estimated value of the frequency offset factor and the estimated value of the instantaneous signal-to-noise ratio.

[0093] On the other hand, you can also use the formula The second algorithm weight is calculated, where is the first algorithm weight, Is a logical judgment symbol, if If established The value is 1, otherwise it is 0. Represents the decision sample space Medium sample The algorithm label in the corresponding label point is the second algorithm.

[0094] That is to say, when calculating the weight of the second algorithm, we can first select all the label points whose algorithm labels are the second algorithm in the decision sample space, and then calculate the Euclidean distance and non-zero minimum value of each label point from the label point to the decision point. Finally, add up the reciprocals of all the label points to get the weight of the second algorithm.

[0095] After the first algorithm weight and the second algorithm weight are determined, a target weight may be determined based on the first algorithm weight and the second algorithm weight.

[0096] In one example, if the first algorithm weight is greater than or equal to the second algorithm weight, the target algorithm can be determined to be the first algorithm; conversely, if the second weight is greater than the first weight, the target algorithm can be determined to be the second algorithm.

[0097] If the target algorithm is determined to be the first algorithm, the Doppler frequency compensation value of the pilot signal can be determined in the following manner: first, obtain the peak value of the pilot signal spectrum. and Two adjacent spectral lines and , then use the formula Determine the Doppler frequency compensation value of the pilot signal, wherein: is the Doppler frequency compensation value calculated using the first algorithm, is the sampling rate for FFT transformation of each signal segment, is the number of points for FFT transformation of each signal segment, is the frequency estimation direction, The value is 1 or -1, is the estimated value of the frequency offset factor.

[0098] On the other hand, if the target algorithm is determined to be the second algorithm, the Doppler frequency compensation value of the pilot signal can be determined in the following manner: first, the pilot signal is down-converted using the rough estimate of the Doppler frequency of the pilot signal to obtain a low-frequency signal. ,in is a low-frequency signal, is the pilot signal, is the rough estimate of the Doppler frequency of the pilot, is the sampling rate of FFT transformation for each signal segment, n is the time domain sampling point index of the pilot signal, Is the imaginary number symbol.

[0099] Then, a frequency search is performed on the low-frequency signal near the zero frequency to determine the frequency value that minimizes the likelihood function and the Doppler frequency compensation value of the pilot. for:

[0100]

[0101] in, The independent variable is The function is The minimum independent variable point set function of is the likelihood function of the low-frequency signal, is the number of points of the low-frequency signal.

[0102] After determining the Doppler frequency compensation value of the pilot, the sum of the coarse Doppler frequency estimation value of the pilot and the Doppler frequency compensation value of the pilot can be determined as the precise Doppler frequency value of the pilot, thereby completing the Doppler estimation of the low-orbit opportunity signal.

[0103] Figure 9 This is a flow chart of another method for estimating the Doppler of low-orbit opportunity signals for dynamic multi-beam switching provided by the present invention. Figure 9 As shown, the method includes the following steps: first, performing segmented FFT processing on the received LEO signal of opportunity to achieve a coarse Doppler frequency estimate of the pilot signal; then, combining locally generated a priori independent word information, the starting position of the pilot signal is accurately located to remove the pilot signal. Feature parameters are then extracted from the stripped pilot signal. The extracted feature parameters are input into a pre-trained N-decision model. If the decision is Rife, the Rife algorithm is used to perform a fine Doppler frequency estimate on the pilot signal; if the decision is Multi-Less Equivalence (MLE), the MLE algorithm is used to perform a fine Doppler frequency estimate on the pilot signal. Finally, the Doppler frequency estimate of the LEO signal of opportunity is obtained by combining the coarse and fine Doppler frequency estimates of the pilot signal.

[0104] The technical solution provided by the embodiment of the present invention can achieve high-precision robust estimation in large dynamic Doppler scenarios. The estimation accuracy of traditional frequency refinement estimation algorithms is constrained by large parameter variation conditions in implementation. For example, the performance of the Rife algorithm deteriorates or even fails under conditions of small frequency offset factors, and the estimation accuracy of the MLE drops significantly under conditions of low signal-to-noise ratio. The technical solution provided by the embodiment of the present invention adaptively selects the optimal precise frequency estimation algorithm for the low-orbit signal-of-opportunity frame under different parameters based on the extracted characteristic parameters of each low-orbit signal-of-opportunity frame, thereby achieving complementary advantageous parameter intervals between different frequency refinement algorithms, and can obtain higher estimation accuracy under conditions of a wide range of Doppler frequency and signal-to-noise ratio fluctuations of the measured low-orbit signal-of-opportunity signal.

[0105] Furthermore, the technical solution provided by the embodiments of the present invention can accurately identify the location of pilot signals within the complex context of low-orbit signals of opportunity (LEOs) and extract characteristic parameters such as their frequency offset factor and signal-to-noise ratio. By establishing a database of characteristic parameter-algorithm mappings, the optimal path for the algorithm refinement can be adaptively matched to achieve optimal estimator matching for different signal characteristics.

[0106] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A Doppler estimation method for low-orbit signal of opportunity with dynamic multi-beam switching, characterized in that: The method comprises: Step 1: Divide the received LEO opportunity signal into blocks according to a preset period, and pre-process each block signal to obtain a target time frame signal, wherein the target time frame signal includes at least a pilot and a unique word of the LEO opportunity signal; wherein the preset period matches the beam polling period of the LEO opportunity signal; Step 2: Segment the target time frame signal, and perform fast Fourier transform (FFT) on each segment signal of the target time frame signal to obtain the frequency domain peak value of each segment signal, determine the segment signal whose frequency domain peak value meets the preset peak value condition as the target segment signal, and determine the Doppler frequency rough estimation value of the pilot signal based on the frequency domain peak value in the target segment signal; wherein, the preset peak value condition includes that the frequency domain peak value of the current segment signal is greater than a preset peak value threshold, and after sorting the frequency domain peak values ​​of all segment signals from large to small, the frequency domain peak value of the current segment signal is located in the first N digits, where N is a positive integer; Step 3: Generate a local independent word signal using prior information, determine the peak of the cross-correlation function between the local independent word signal and the panned pilot coarse position signal as the pilot start position, and the panned pilot coarse position signal is a signal whose distance from the pilot coarse position is less than a preset distance threshold; wherein the prior information includes at least prior pilot information and prior unique word information; Step 4: Determine a pilot signal based on the pilot starting position, perform FFT processing on the pilot signal again to obtain a pilot signal spectrum, and determine a frequency offset factor estimate and an instantaneous signal-to-noise ratio estimate based on the pilot signal spectrum; wherein the number of points in the FFT processing again is the same as the number of time-domain sampling points of the pilot signal; Step 5: Input the frequency offset factor estimate and the instantaneous signal-to-noise ratio estimate into a pre-trained nearest neighbor node algorithm KNN decision model to obtain a first algorithm weight and a second algorithm weight; wherein the first algorithm is a frequency estimation Rife algorithm based on discrete-time signal phase difference, and the second algorithm is a maximum likelihood estimation MLE algorithm; Step 6: Use the target algorithm to determine the Doppler frequency compensation value of the pilot, and determine the precise value of the Doppler frequency of the pilot based on the rough estimate of the Doppler frequency of the pilot and the Doppler frequency compensation value of the pilot; wherein, when the weight of the first algorithm is greater than the weight of the second algorithm, the target algorithm is the first algorithm; when the weight of the second algorithm is greater than the weight of the first algorithm, the target algorithm is the second algorithm.

2. The method for Doppler estimation of low-orbit signal of opportunity for multi-beam dynamic switching according to claim 1, characterized in that: The step 1 comprises: Dividing the received low-orbit opportunity signal into block signals of fixed duration according to the beam polling period of the low-orbit opportunity signal; Bandpass filter each block signal to obtain the target time frame signal.

3. The method for Doppler estimation of low-orbit signal of opportunity for multi-beam dynamic switching according to claim 1, characterized in that: The step 2 includes: Each block signal is divided into multiple equal-length signal segments, and each signal segment is independently subjected to FFT transformation to obtain the frequency domain peak value of each segment signal; Determine the target segmented signal and obtain the frequency domain peak value of each target segmented signal; Using the formula Determine a rough estimate of the Doppler frequency of the pilot based on the rough position of the pilot; wherein, is the rough estimate of the Doppler frequency of the pilot, is the spectral line number corresponding to the maximum peak value of each target segment signal, is the sampling rate for FFT transformation of each signal segment, is the number of points used to perform FFT transformation on each signal segment.

4. The method for Doppler estimation of low-orbit signal of opportunity for multi-beam dynamic switching according to claim 1, characterized in that: The step 3 comprises: Generate local independent word signals using prior information ,in is the local independent word signal, is the a priori pilot amplitude, is the a priori unique word amplitude, is the imaginary number symbol, is the rough estimate of the Doppler frequency of the pilot, is the sampling rate for FFT transformation of each signal segment, n is the time domain sampling point index of the pilot signal, and each n value represents a sampling point. is the number of sampling points for each independent symbol of each LEO signal of opportunity, is the a priori unique character symbol information, is the rounding symbol; Construct a cross-correlation function between the local independent word signal and the pan-pilot coarse position signal, and determine the peak of the cross-correlation function as the pilot starting position ; The pilot starting position is ,in The function with the independent variable k is The maximum independent variable point set function of is the cross-correlation function, is the point index of the cross-correlation function, for The number of points, is the rough position signal of the rich pilot.

5. The method for Doppler estimation of low-orbit signal of opportunity for multi-beam dynamic switching according to claim 1, characterized in that: The step 4 comprises: Determine the peak value of the pilot signal spectrum ; Extract the spectrum peak Two adjacent spectral lines and ; Using the formula: ； ； The frequency offset factor estimate is calculated, where is the estimated value of the frequency offset factor, is the frequency estimation direction.

6. The method for Doppler estimation of low-orbit signal of opportunity for multi-beam dynamic switching according to claim 1, characterized in that: The step 4 further comprises: Determine the peak value of the pilot signal spectrum ; Using the formula The instantaneous signal-to-noise ratio estimation value is calculated, wherein is the instantaneous signal-to-noise ratio estimate, is the number of points of the FFT processing performed again, is the spectral line index of the pilot signal, is the first spectral lines.

7. The method for Doppler estimation of low-orbit signal of opportunity for multi-beam dynamic switching according to claim 1, characterized in that: The pre-trained KNN decision model in step 5 is trained in the following way: Monte Carlo simulation is performed within a preset frequency offset factor and a preset instantaneous signal-to-noise ratio range based on real Iridium satellite signal statistics to obtain the root mean square error of the Doppler frequency estimation accuracy when the first algorithm and the second algorithm are used for different combinations of frequency offset factors and instantaneous signal-to-noise ratios; Determining an algorithm label for each frequency offset factor and instantaneous signal-to-noise ratio combination based on the precision error to obtain a label point in the pre-trained KNN decision model; Each tag point includes a pair of frequency offset factor and instantaneous signal-to-noise ratio combination, and an algorithm tag of the tag point, where the algorithm tag is the first algorithm or the second algorithm.

8. The method for Doppler estimation of low-orbit signal of opportunity for multi-beam dynamic switching according to claim 7, characterized in that: The step 5 comprises: Inputting the frequency offset factor estimation value and the instantaneous signal-to-noise ratio estimation value into the pre-trained decision model, and calculating the Euclidean distance between the combination of the frequency offset factor estimation value and the instantaneous signal-to-noise ratio estimation value and each label point; Select K label points with the smallest Euclidean distance as the decision sample space ,in, is the decision sample space, K is a positive integer; Using the formula The weight of the first algorithm is calculated, where is the weight of the first algorithm, Is a logical judgment symbol, if If established The value is 1, otherwise it is 0. Represents the decision sample space Medium sample The algorithm label in the corresponding label point is the first algorithm, is a non-zero decimal number used to avoid the denominator of a fraction being zero; Using the formula The second algorithm weight is calculated, where is the weight of the first algorithm, Is a logical judgment symbol, if If established The value is 1, otherwise it is 0. Represents the decision sample space Medium sample The algorithm label in the corresponding label point is the second algorithm.

9. The method for Doppler estimation of low-orbit signal of opportunity for multi-beam dynamic switching according to claim 1, characterized in that: The step 6 comprises: In response to determining that the target algorithm is the first algorithm, obtaining the pilot signal spectrum peak and Two adjacent spectral lines and ; Using the formula Determine the Doppler frequency compensation value of the pilot signal, wherein: is the Doppler frequency compensation value calculated using the first algorithm, is the sampling rate for FFT transformation of each signal segment, is the number of points for FFT transformation of each signal segment, is the frequency estimation direction, The value is 1 or -1, is the estimated value of the frequency offset factor; The sum of the rough estimation value of the Doppler frequency of the pilot and the Doppler frequency compensation value of the pilot is determined to be the precise value of the Doppler frequency of the pilot.

10. The method for Doppler estimation of low-orbit signal of opportunity for multi-beam dynamic switching according to claim 1, characterized in that: The step 6 further comprises: In response to determining that the target algorithm is the second algorithm, the pilot signal is down-converted using the rough Doppler frequency estimate of the pilot signal to obtain a low-frequency signal ,in is the low-frequency signal, is the pilot signal, is a rough estimate of the Doppler frequency of the pilot signal, is the sampling rate of FFT transformation for each signal segment, n is the time domain sampling point index of the pilot signal, is the imaginary number symbol; Perform a frequency search on the low-frequency signal near the zero frequency to determine the frequency value that minimizes the likelihood function and take it as the Doppler frequency compensation value of the pilot. The Doppler frequency compensation value of the pilot is for: ； in, The independent variable is The function is The minimum independent variable point set function of is the likelihood function of the low-frequency signal, is the number of points of the low-frequency signal; The sum of the rough estimation value of the Doppler frequency of the pilot and the Doppler frequency compensation value of the pilot is determined to be the precise value of the Doppler frequency of the pilot.

Images