A method and system for detecting Starlink terminal wireless signals.
By performing multi-channel intermediate frequency sampling, carrier frequency correction, and ephemeris visible window gating on the Starlink terminal wireless signal, combined with Doppler inversion and angular frequency consistency verification, the problem of insufficient candidate signal consistency verification in Starlink terminal detection was solved, and highly reliable Starlink terminal detection was achieved.
Patent Information
- Application Number
- CN202611624305.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-20
- Publication Date
- 2026-11-17
AI Technical Summary
In existing technologies, the consistency verification between candidate signals and ephemeris motion is insufficient in Starlink terminal wireless signal detection, resulting in insufficient terminal detection reliability and difficulty in verifying angular frequency evidence.
By acquiring multi-channel intermediate frequency sampling sequences, synchronous sampling alignment and carrier frequency correction are performed to form time-frequency burst observation frames. Combined with ephemeris visibility windows and Doppler drift gating, non-ephemeris consistency interference is eliminated to form terminal uplink candidate clusters. The carrier frequency trajectory is continuously tracked and Doppler inversion registration is performed to extract the pre-compensated residual spectrum. Segmented Kalman smoothing and abrupt change removal are performed to form the pre-compensated inversion trajectory. The array phase difference sequence and angle of arrival estimation sequence are extracted synchronously and the angular frequency consistency closure check is performed to form the Starlink terminal closure evidence chain.
It achieves the separation of terminal uplink transmission characteristics from low-orbit satellite motion factors, and has the capabilities of continuous tracking, spatial positioning, and status verification, thereby improving the reliability and accuracy of Starlink terminal detection.
Smart Images

Figure CN123122640A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of satellite communication signal detection technology, and in particular to a method and system for detecting wireless signals from Starlink terminals. Background Technology
[0002] Satellite internet terminals rely on low-Earth orbit (LEO) satellite constellations for uplink access, playing a crucial role in regional radio monitoring, spectrum order maintenance, and space communication situational awareness. Conventional methods for detecting Starlink terminal wireless signals typically involve wideband reception, time-frequency energy analysis, burst signal extraction, and frequency trajectory tracking within the target satellite's communication frequency band. These methods, combined with the array reception capabilities of monitoring stations, allow for direction estimation and continuous observation of suspected uplink signals, thus forming the foundation for detecting wireless signals from LEO satellite terminal transmission activities.
[0003] In the above-mentioned conventional detection methods, on the one hand, low-orbit satellites move at high speed relative to the monitoring point, and the uplink signal of the terminal is significantly affected by ephemeris visibility and Doppler drift. If only time-frequency burst features are used for screening, narrowband burst interference that is not consistent with ephemeris is easily included in the candidate. On the other hand, when there is a lack of closed verification between the terminal uplink pre-compensation features, satellite motion direction, burst boundary continuity and angle of arrival stability, it is difficult to form a stable terminal detection chain with single frequency evidence or single angle evidence. Summary of the Invention
[0004] In view of the problems existing in the prior art, the present invention is proposed.
[0005] Therefore, this invention provides a method and system for detecting Starlink terminal wireless signals, which solves the problem of insufficient verification of the consistency between candidate signals and ephemeris motion and the difficulty in closing the verification of angular frequency evidence in existing Starlink terminal wireless signal detection, resulting in insufficient reliability of terminal detection.
[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for detecting wireless signals of Starlink terminals, comprising: acquiring multi-channel intermediate frequency (IF) sampling sequences of the target satellite communication frequency band within a monitoring area; performing synchronous sampling alignment, carrier frequency correction, and array channel amplitude and phase calibration on the multi-channel IF sampling sequences; and performing short-time power spectrum estimation, burst envelope segmentation, and carrier frequency trajectory tracking through a time-frequency analysis window to form a time-frequency burst observation frame; calculating the Starlink ephemeris visibility window according to the monitoring point location, monitoring time, and low-Earth orbit satellite orbit information; performing ephemeris time gating and Doppler drift gating on the time-frequency burst observation frame and the Starlink ephemeris visibility window to eliminate non-ephemeris consistent interference and form a terminal uplink candidate cluster; continuously tracking the carrier frequency trajectory in the terminal uplink candidate cluster; and calculating the ephemeris Doppler trajectory according to the candidate visible satellite orbit trajectory. Doppler inversion registration is performed between the carrier frequency trajectory and the ephemeris Doppler trajectory. The uplink pre-compensated residual spectrum is extracted, and segmented Kalman smoothing and abrupt elimination are performed on the uplink pre-compensated residual spectrum according to the burst envelope boundary to form the pre-compensated inversion trajectory. The array phase difference sequence, array amplitude difference sequence, and angle of arrival estimation sequence are extracted simultaneously from the pre-compensated inversion trajectory. The angle of arrival estimation sequence is projected onto the geographic grid of the monitoring area, and then the angular frequency consistency closure check is performed according to the uplink pre-compensated residual spectrum, the motion direction of candidate visible satellites, the continuity of burst envelope boundary, and the stability of angle of arrival to form the Starlink terminal closure evidence chain. The Starlink terminal closure evidence chain is time-series condensed in chronological order, and the finite state transition is verified for adjacent Starlink terminal closure evidence chains. The Starlink terminal detection evidence index is compiled, and the Starlink terminal wireless signal detection record is output.
[0007] As a preferred embodiment of the method for detecting Starlink terminal wireless signals according to the present invention, the step of forming time-frequency burst observation frames is specifically as follows: Within the monitoring area, the target satellite communication frequency band is received synchronously through multiple channels. The radio frequency signals of each receiving channel are converted into intermediate frequency sampling signals. The intermediate frequency sampling signals are then uniformly organized according to the sampling time and channel identifier to form a multi-channel intermediate frequency sampling sequence. The multi-channel intermediate frequency sampling sequence is time-aligned between channels according to the sampling time. The carrier frequency of the time-aligned multi-channel intermediate frequency sampling sequence is corrected. The inter-channel amplitude difference correction and inter-channel phase difference correction are performed on the carrier frequency corrected multi-channel intermediate frequency sampling sequence with the reference channel as the reference, forming an array amplitude and phase calibration sequence. The array amplitude and phase calibration sequence is divided into continuous time-frequency analysis windows. Short-time power spectrum estimation is performed on the continuous time-frequency analysis windows. Energy mutation regions in the short-time power spectrum estimation are extracted along the observation time. The energy mutation regions are divided into burst envelope regions. The time continuous path of the center frequency point is tracked within the burst envelope region to form a time-frequency burst observation frame.
[0008] As a preferred embodiment of the method for detecting Starlink terminal wireless signals described in this invention, the step of performing ephemeris-time gating and Doppler drift gating on the time-frequency burst observation frames and the Starlink ephemeris visible window is as follows: The monitoring point location is converted into ground observation coordinates, and the low-orbit satellite orbit information is extrapolated according to the monitoring time to form a transit trajectory. The elevation angle entry time, elevation angle departure time, azimuth angle sweep interval and radial velocity change interval of candidate visible satellites are extracted along the transit trajectory to form the Starlink ephemeris visible window. The burst envelope region and carrier frequency trajectory are extracted from the time-frequency burst observation frame. The start and end times of the burst envelope region are overlapped with the elevation entry time and elevation exit time in the Starlink ephemeris visible window for screening. The corresponding carrier frequency trajectory is retained for the burst envelope region that falls into the Starlink ephemeris visible window, and the corresponding carrier frequency trajectory is removed for the burst envelope region that does not fall into the Starlink ephemeris visible window, thus forming the ephemeris time-gated observation frame. Based on the radial velocity variation range in the Starlink ephemeris visible window, the Doppler drift direction and Doppler drift amplitude range corresponding to the candidate visible satellites are calculated. The measured frequency drift direction and measured frequency drift amplitude are extracted along the carrier frequency trajectory in the ephemeris time-gated observation frame. The measured frequency drift direction is compared with the Doppler drift direction for directional consistency, and the measured frequency drift amplitude is compared with the Doppler drift amplitude range for amplitude fall area comparison, thus forming the Doppler drift-gated observation frame.
[0009] As a preferred embodiment of the method for detecting Starlink terminal wireless signals according to the present invention, the step of forming a terminal uplink candidate cluster is specifically as follows: For each burst object in the Doppler drift-gated observation frame, the Starlink ephemeris visibility window, measured frequency drift direction, measured frequency drift amplitude, and burst duration are checked. Burst objects that do not meet the Starlink ephemeris visibility window requirement are removed by time, burst objects that do not meet the measured frequency drift direction requirement are removed by drift, burst objects that do not meet the measured frequency drift amplitude requirement are removed by amplitude, and burst objects that do not meet the burst duration continuity requirement are removed by continuity, thus forming a set of burst objects with consistent ephemeris. The set of burst objects with consistent ephemeris is merged according to observation time, frequency band location and candidate visible satellites. Burst objects within the same transit time period are merged according to the adjacency of observation time, burst objects within the same frequency band are merged according to the adjacency of frequency band location, and burst objects under the same candidate visible satellite are merged according to the affiliation of candidate visible satellites, forming a terminal uplink candidate cluster.
[0010] As a preferred embodiment of the method for detecting Starlink terminal wireless signals described in this invention, the step of continuously tracking the carrier frequency trajectory in the uplink candidate cluster of the terminal and calculating the ephemeris Doppler trajectory according to the orbital trajectory of the candidate visible satellites is as follows: Based on the burst objects in the terminal uplink candidate cluster, the carrier frequency trajectory corresponding to each burst object is extracted along the observation time. The trajectory is then continued according to the frequency band proximity relationship and the consistent frequency drift direction relationship between adjacent burst objects to form a continuous carrier frequency trajectory. The observation time corresponding to the continuous carrier frequency trajectory is mapped to the orbital trajectory of the candidate visible satellite. The radial velocity change of the candidate visible satellite relative to the position of the monitoring point during the observation time is extracted, and the frequency offset change corresponding to the radial velocity change is calculated according to the communication frequency band of the target satellite to form the ephemeris Doppler trajectory.
[0011] As a preferred embodiment of the method for detecting Starlink terminal wireless signals according to the present invention, the step of forming a pre-compensated inversion trajectory is specifically as follows: The continuous carrier frequency trajectory and the ephemeris Doppler trajectory are aligned along the time axis according to the observation time. The ephemeris Doppler trajectory is superimposed onto the continuous carrier frequency trajectory according to the reverse compensation relationship. The frequency drift component caused by the motion of candidate visible satellites in the continuous carrier frequency trajectory is inverted and subtracted to form the inverted registration frequency trajectory. Residual sampling is performed on the inverted registration frequency trajectory along the observation time. The residual sampled values are mapped to the time-frequency position of the corresponding burst object according to the burst envelope boundary. The frequency distribution statistics of the residual sampled values within the same burst object are performed to form the uplink pre-compensation residual spectrum. The uplink pre-compensated residual spectrum is divided into multiple residual spectrum segments according to the burst envelope boundary. Within each residual spectrum segment, the residual state is recursively deduced according to the observation time. Kalman smoothing is performed on the residual fluctuations between adjacent observation times to form a segmented smoothed residual spectrum. The variation amplitude of adjacent residual states in the piecewise smooth residual spectrum is compared along the observation time. Residual states that deviate from the continuity relationship of the burst envelope boundary and the continuity relationship of the carrier frequency trajectory are eliminated. The remaining residual states are reconnected according to the observation time to form a pre-compensated inversion trajectory.
[0012] As a preferred embodiment of the method for detecting Starlink terminal wireless signals described in this invention, the step of simultaneously extracting the array phase difference sequence, array amplitude difference sequence, and angle of arrival estimation sequence from the pre-compensated inversion trajectory, and projecting the angle of arrival estimation sequence onto the geographic grid of the monitoring area, is as follows: Using the observation time and burst envelope boundary in the pre-compensated inversion trajectory as indexes, array received data of the same burst object are extracted from the multi-channel intermediate frequency sampling sequence corresponding to the pre-compensated inversion trajectory. The phase differences between each array received channel at the same observation time are arranged to form an array phase difference sequence, and the amplitude differences between each array received channel at the same observation time are arranged to form an array amplitude difference sequence. The angle of arrival is estimated based on the array phase difference sequence and the array amplitude difference sequence to form an angle of arrival estimation sequence. Using the direction of the angle of arrival in the angle of arrival estimation sequence as the projection direction and the grid position in the geographic grid of the monitoring area as the projection object, the angle of arrival direction and the grid position in the geographic grid of the monitoring area are intersected angularly, and the observation time and burst envelope boundary corresponding to the angular intersection position are marked on the geographic grid of the monitoring area to form the angle of arrival grid projection map.
[0013] As a preferred embodiment of the method for detecting Starlink terminal wireless signals according to the present invention, the step of forming a closed evidence chain for the Starlink terminal is specifically as follows: The uplink pre-compensated residual spectrum corresponding to the angle of arrival grid projection map is time-aligned with the motion direction of the candidate visible satellite. The residual change direction of the uplink pre-compensated residual spectrum and the Doppler change direction corresponding to the motion direction of the candidate visible satellite are compared along the observation time. The observation times with matching directions are written into frequency closure markers to form a frequency closure marker sequence. The continuity of the burst envelope boundary corresponding to the frequency closure mark sequence is time-aligned with the angle of arrival grid projection map. The continuity relationship of the burst envelope boundary and the angle fluctuation range of the angle of arrival estimation sequence are checked along adjacent burst objects. Burst objects with continuous continuity relationship and stable angle fluctuation range are written into spatial closure marks to form a spatial closure mark sequence. The frequency closure marker sequence and the spatial closure marker sequence are merged in the same location according to the observation time, the burst envelope boundary and the geographical grid of the monitoring area. The uplink pre-compensation residual spectrum, the motion direction of candidate visible satellites, the continuity of the burst envelope boundary and the stability of the angle of arrival in the same burst object are connected in a closure relationship to form a Starlink terminal closure evidence chain.
[0014] As a preferred embodiment of the Starlink terminal wireless signal detection method of the present invention, the output of the Starlink terminal wireless signal detection record is as follows: Using the observation time in the Starlink terminal closure evidence chain as the primary order and the candidate visible satellite, frequency band location, and monitoring area geographic grid as the secondary order, Starlink terminal closure evidence chains that belong to the same candidate visible satellite and have continuous observation time are sequentially arranged and adjacently labeled to form adjacent Starlink terminal closure evidence chain groups. According to the frequency closure marker sequence, spatial closure marker sequence, burst envelope boundary continuity relationship, and uplink pre-compensation residual spectrum continuity relationship in the adjacent Starlink terminal closure evidence chain group, the acquisition state, stable transmission state, beam switching state, and disappearance state are marked for each Starlink terminal closure evidence chain. The transition order between adjacent states is checked according to the observation time to form a finite state transition sequence. The finite state transition sequence is mapped to adjacent Starlink terminal closure evidence chain groups according to the observation time. For each Starlink terminal closure evidence chain, candidate visible satellites, frequency band positions, monitoring area geographic grids, frequency closure marker sequences, spatial closure marker sequences, and status labels are recorded to form a Starlink terminal detection evidence index. The detection time, detection frequency band, candidate visible satellites, terminal transmission status, detection confidence level, and alarm level are organized according to the status labels, frequency closure marker sequences, and spatial closure marker sequences in the Starlink terminal detection evidence index to form a Starlink terminal wireless signal detection record.
[0015] Secondly, this invention provides a detection system for Starlink terminal wireless signals, comprising: a sampling observation module, used to collect multi-channel intermediate frequency (IF) sampling sequences of the target satellite communication frequency band within the monitoring area, perform synchronous sampling alignment, carrier frequency correction, and array channel amplitude and phase calibration on the multi-channel IF sampling sequences, and perform short-time power spectrum estimation, burst envelope segmentation, and carrier frequency trajectory tracking through a time-frequency analysis window to form a time-frequency burst observation frame; a star ephemeris gating module, used to calculate the Starlink star ephemeris visible window according to the monitoring point location, monitoring time, and low-Earth orbit satellite orbit information, and perform star ephemeris time gating and Doppler drift gating on the time-frequency burst observation frame and the Starlink star ephemeris visible window to eliminate non-ephemeris consistent interference and form a terminal uplink candidate cluster; and a residual inversion module, used to continuously track the carrier frequency trajectory in the terminal uplink candidate cluster and calculate the star ephemeris Doppler according to the candidate visible satellite orbit trajectory. The trajectory module performs Doppler inversion registration between the carrier frequency trajectory and the ephemeris Doppler trajectory, extracts the uplink pre-compensated residual spectrum, and performs segmented Kalman smoothing and abrupt elimination on the uplink pre-compensated residual spectrum according to the burst envelope boundary to form the pre-compensated inversion trajectory. The angular frequency closure module is used to simultaneously extract the array phase difference sequence, array amplitude difference sequence, and angle of arrival estimation sequence from the pre-compensated inversion trajectory. The angle of arrival estimation sequence is projected onto the geographic grid of the monitoring area, and then the angular frequency consistency closure verification is performed according to the uplink pre-compensated residual spectrum, the motion direction of candidate visible satellites, the continuity of burst envelope boundaries, and the stability of angle of arrival to form the Starlink terminal closure evidence chain. The evidence index module is used to perform temporal aggregation of the Starlink terminal closure evidence chain in chronological order, perform finite state transition verification on adjacent Starlink terminal closure evidence chains, compile the Starlink terminal detection evidence index, and output the Starlink terminal wireless signal detection record.
[0016] The beneficial effects of this invention are as follows: by continuously tracking the carrier frequency trajectory in the uplink candidate cluster of the terminal and performing Doppler inversion registration with the ephemeris Doppler trajectory, the uplink pre-compensated residual spectrum is extracted and a pre-compensated inversion trajectory is formed, thereby achieving the separation of the terminal's uplink transmission characteristics from the motion factors of the low-orbit satellite; then, by synchronously extracting the array phase difference sequence, array amplitude difference sequence, and angle of arrival estimation sequence from the pre-compensated inversion trajectory and performing angular frequency consistency closure verification, a Starlink terminal closed evidence chain is formed, which is used for compiling the Starlink terminal detection evidence index and outputting the Starlink terminal wireless signal detection record, enabling the detection chain to have continuous tracking, spatial positioning, and status verification capabilities. Attached Figure Description
[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 This is a flowchart of a method for detecting wireless signals from Starlink terminals.
[0019] Figure 2 This is a schematic diagram of a system for detecting wireless signals from Starlink terminals.
[0020] Figure 3 The graph shows the test results of F1 and relative improvement under different signal-to-noise ratio groups.
[0021] Figure 4 This is a graph showing the results of standardized performance profile tests for multiple indicators. Detailed Implementation
[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.
[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.
[0025] Example 1: Reference Figures 1-4 This is one embodiment of the present invention, which provides a method for detecting Starlink terminal wireless signals, including the following steps: S1: Collect multi-channel intermediate frequency sampling sequences of the target satellite communication frequency band within the monitoring area, perform synchronous sampling alignment, carrier frequency correction and array channel amplitude and phase calibration on the multi-channel intermediate frequency sampling sequences, and perform short-time power spectrum estimation, burst envelope segmentation and carrier frequency trajectory tracking through the time-frequency analysis window to form a time-frequency burst observation frame; S1.1: Within the monitoring area, the target satellite communication frequency band is received synchronously through multiple channels. The radio frequency signals of each receiving channel are converted into intermediate frequency sampling signals. The intermediate frequency sampling signals are then uniformly organized according to the sampling time and channel identifier to form a multi-channel intermediate frequency sampling sequence. Furthermore, within the monitoring area, the target satellite communication frequency band is received synchronously through multiple channels. Multiple receiving channels use the same reference clock and the same sampling trigger signal to sample the radio frequency signal in the target satellite communication frequency band in parallel. Each receiving channel converts the radio frequency signal into an intermediate frequency sampling signal through down-conversion processing. The intermediate frequency (IF) sampling signals are labeled with sampling time, channel identifier, and center frequency. Then, the IF sampling signals are uniformly organized according to the order of sampling time and the correspondence of channel identifiers to form a multi-channel IF sampling sequence.
[0026] S1.2: Perform inter-channel time alignment on the multi-channel intermediate frequency sampling sequence according to the sampling time, perform carrier frequency correction on the inter-channel time-aligned multi-channel intermediate frequency sampling sequence, and perform inter-channel amplitude difference correction and inter-channel phase difference correction on the carrier frequency-corrected multi-channel intermediate frequency sampling sequence with the reference channel as the reference, to form an array amplitude and phase calibration sequence. Furthermore, the multi-channel intermediate frequency (IF) sampling sequence is time-aligned between channels according to the sampling time. The IF sampling signal of one receiving channel is selected as the IF sampling signal of the reference channel. Within the same sampling time range, the IF sampling signal of the reference channel and the IF sampling signals of other receiving channels are extracted. Cross-correlation is calculated between the IF sampling signal of the reference channel and the IF sampling signal of other receiving channels. The sampling time deviation is obtained based on the integer sampling point where the cross-correlation peak is located and the interpolation position of the neighborhood of the cross-correlation peak. The multi-channel IF sampling sequence is then time-rearranged and interpolated and resampled according to the sampling time deviation. The multi-channel intermediate frequency sampling sequence after inter-channel time alignment is divided into frequency correction windows. The actual center frequency point corresponding to the energy peak within the frequency correction window is extracted. The difference between the actual center frequency point and the set intermediate frequency center frequency point corresponding to the target satellite communication frequency band is recorded as the frequency offset. Digital frequency shift compensation is performed according to the frequency offset. Before receiving the target satellite communication frequency band, the same source calibration signal is allocated to each receiving channel. The fixed amplitude response difference and fixed phase response difference of each receiving channel relative to the reference channel are calculated to obtain the channel amplitude and phase calibration coefficient. Then, the fixed amplitude response difference compensation and fixed phase response difference compensation are performed on the multi-channel intermediate frequency sampling sequence after carrier frequency correction according to the channel amplitude and phase calibration coefficient. The spatial phase difference formed by the radio frequency signal in the target satellite communication frequency band arriving at each receiving channel is preserved to form the array amplitude and phase calibration sequence.
[0027] S1.3: Divide the array amplitude and phase calibration sequence into a continuous time-frequency analysis window, perform short-time power spectrum estimation on the continuous time-frequency analysis window, extract the energy change region in the short-time power spectrum estimation along the observation time, divide the energy change region into a burst envelope region, and track the time continuous path of the center frequency point within the burst envelope region to form a time-frequency burst observation frame; Furthermore, the array amplitude and phase calibration sequence is divided into continuous time-frequency analysis windows according to the observation time, and an overlapping sampling interval is set between adjacent continuous time-frequency analysis windows. Frequency domain energy calculation is performed on the array amplitude and phase calibration sequence in each continuous time-frequency analysis window to form a short-time power spectrum estimate. Energy changes in the same frequency band in short-time power spectrum estimation are calculated along the observation time. The median statistical value of the energy of non-burst frequency bands within the continuous time-frequency analysis window is used as the noise floor, and the range of received noise fluctuations is used as the energy margin. The time-frequency region with energy higher than the sum of the noise floor and the energy margin and continuously spanning adjacent continuous time-frequency analysis windows is divided into burst envelope regions. Within the burst envelope region, the energy-weighted center frequency is extracted window by window. Energy-weighted center frequencies with frequency differences between adjacent consecutive time-frequency analysis windows that fall within the frequency continuity range are connected. The frequency continuity range is set according to the frequency resolution of the consecutive time-frequency analysis windows and the allowable frequency drift range corresponding to the target satellite communication frequency band. The direction and magnitude of change of the energy-weighted center frequency are recorded along the observation time (the energy-weighted center frequencies connected according to the observation time, as well as the direction and magnitude of change of the energy-weighted center frequencies, are recorded as the carrier frequency trajectory). The short-time power spectrum estimation, burst envelope region, and carrier frequency trajectory are indexed and labeled according to the observation time, frequency band position, and channel identifier to form a time-frequency burst observation frame.
[0028] It should be noted that the inter-channel time alignment uses the integer sampling point of the cross-correlation peak and the interpolation position of the cross-correlation peak neighborhood to obtain the sampling time deviation, avoiding the limitation of only processing integer sampling point offsets; the array channel amplitude and phase calibration uses the same source calibration signal to obtain the channel amplitude and phase calibration coefficients, which only compensates for the fixed amplitude response difference and fixed phase response difference of each receiving channel, and does not eliminate the spatial phase difference formed by the radio frequency signal arriving at each receiving channel within the target satellite communication frequency band. Therefore, the array amplitude and phase calibration sequence can still support the extraction of subsequent array phase difference sequences and angle of arrival estimation sequences; the burst envelope region is divided by noise floor, energy margin and continuous time-frequency analysis window, and the carrier frequency trajectory is connected by the energy weighted center frequency point and the continuous frequency range. The burst envelope region and carrier frequency trajectory in the time-frequency burst observation frame can continue to be used for ephemeris time gating and Doppler drift gating; The allowable frequency drift range corresponding to the target satellite communication frequency band is jointly set based on the frequency drift calculated from the maximum radial velocity change in the low-Earth orbit satellite orbit information, the residual frequency offset of the received local oscillator, and the frequency resolution estimated by the short-time power spectrum.
[0029] S2: Calculate the Starlink ephemeris visible window according to the monitoring point location, monitoring time, and low-Earth orbit satellite orbit information. Perform ephemeris time gating and Doppler drift gating on the time-frequency burst observation frames and the Starlink ephemeris visible window to eliminate non-ephemeris consistent interference and form a terminal uplink candidate cluster. S2.1: Convert the monitoring point location into ground observation coordinates, extrapolate the low-orbit satellite orbit information according to the monitoring time to form a transit trajectory, and extract the elevation angle entry time, elevation angle departure time, azimuth angle sweep interval and radial velocity change interval of candidate visible satellites along the transit trajectory to form the Starlink ephemeris visible window; Furthermore, the monitoring point location is converted into ground observation coordinates, which include longitude, latitude, and altitude. First, the monitoring point location is converted into geocentric and geofixed coordinates, and then ground observation coordinates composed of east, north, and zenith directions are established with the monitoring point location as the coordinate origin. The low-Earth orbit satellite orbit information is divided into multiple orbit calculation times according to the monitoring time. The low-Earth orbit satellite orbit information includes satellite identifier, orbital elements and epoch time. At each orbit calculation time, the low-Earth orbit satellite orbit information is extrapolated to obtain the satellite spatial position and satellite velocity of the candidate visible satellite in the geocentric and geofixed coordinate system. The satellite spatial position and satellite velocity are then converted to ground observation coordinates. The elevation and azimuth angles are calculated based on the line of sight between the satellite's spatial position and the monitoring point's position. The radial velocity is calculated based on the projection of the satellite's velocity onto the line of sight. The starting time when the elevation angle reaches the visible elevation angle threshold is recorded as the elevation angle entry time, and the ending time when the elevation angle falls below the visible elevation angle threshold is recorded as the elevation angle departure time. The azimuth angle change between the elevation angle entry time and the elevation angle departure time is recorded as the azimuth angle sweep interval, and the radial velocity change between the elevation angle entry time and the elevation angle departure time is recorded as the radial velocity change interval, thus forming the Starlink ephemeris visible window.
[0030] It should be noted that the visible elevation angle threshold is set according to the greater of the minimum operating elevation angle of the receiving antenna and the elevation angle of ground obstruction in the monitoring area.
[0031] S2.2: Extract the burst envelope region and carrier frequency trajectory from the time-frequency burst observation frame. Filter the burst envelope region by matching the start and end times with the elevation entry and exit times in the Starlink ephemeris visible window. For burst envelope regions that fall into the Starlink ephemeris visible window, retain the corresponding carrier frequency trajectory. For burst envelope regions that do not fall into the Starlink ephemeris visible window, remove the corresponding carrier frequency trajectory to form an ephemeris time-gated observation frame. Furthermore, the burst envelope region and carrier frequency trajectory are extracted from the time-frequency burst observation frames. The burst envelope region includes the burst start time, burst end time, burst duration, and burst occupied frequency band. The carrier frequency trajectory includes the direction and amplitude of change of the energy-weighted center frequency. The burst start time and burst end time of the burst envelope region are compared with the elevation entry time and elevation exit time in the Starlink ephemeris visible window. The time coincidence comparison is performed by expanding the boundary using ephemeris time margin. The ephemeris time margin is set according to the orbit extrapolation time error and the observation time resolution of the time-frequency burst observation frame. For burst envelope regions where the entire burst start time to burst end time falls within the Starlink ephemeris visible window, the complete burst envelope region and its corresponding carrier frequency trajectory are preserved. For burst envelope regions where the burst start time to burst end time partially falls within the Starlink ephemeris visible window, the overlapping time period falling within the Starlink ephemeris visible window is extracted, and the corresponding carrier frequency trajectory within the overlapping time period is extracted. For burst envelope regions where the overlapping time period is less than the minimum gate duration, the corresponding carrier frequency trajectory is removed. The preserved burst envelope regions, preserved carrier frequency trajectories, and candidate visible satellites are labeled according to the observation time to form ephemeris time-gated observation frames.
[0032] It should be noted that the minimum gating duration is set jointly based on the window length of the continuous time-frequency analysis window, the overlapping sampling interval of adjacent continuous time-frequency analysis windows, and the minimum effective duration of the burst envelope region.
[0033] S2.3: Based on the radial velocity variation range in the Starlink ephemeris visible window, calculate the Doppler drift direction and Doppler drift amplitude range corresponding to the candidate visible satellites. Extract the measured frequency drift direction and measured frequency drift amplitude along the carrier frequency trajectory in the ephemeris time-gated observation frame. Compare the measured frequency drift direction with the Doppler drift direction for directional consistency and compare the measured frequency drift amplitude with the Doppler drift amplitude range for amplitude fall area comparison to form a Doppler drift-gated observation frame. Furthermore, based on the radial velocity variation range in the Starlink ephemeris visible window and the target satellite communication frequency band, the Doppler drift direction and Doppler drift amplitude range corresponding to the candidate visible satellite are calculated. The radial velocity variation range is time-sampled according to the observation time in the ephemeris time-gated observation frame, and the radial velocity corresponding to each observation time is converted into ephemeris frequency offset according to the target satellite communication frequency band. The ephemeris frequency offset is connected along the observation time to form the ephemeris frequency offset curve corresponding to the candidate visible satellite. The direction of frequency change of the ephemeris frequency offset curve from the burst start time to the burst end time is recorded as the Doppler drift direction, and the range of frequency change of the ephemeris frequency offset curve from the burst start time to the burst end time time is recorded as the Doppler drift amplitude range; the frequency point change between adjacent observation times is extracted along the carrier frequency trajectory in the ephemeris time-gated observation frame, and the direction of frequency point change with observation time is recorded as the measured frequency drift direction, and the range of frequency point change of the carrier frequency trajectory from the burst start time to the burst end time time is recorded as the measured frequency drift amplitude; The measured frequency drift direction is compared with the Doppler drift direction for directional consistency. The measured frequency drift amplitude is compared with the Doppler drift amplitude range and the frequency gating margin for amplitude fall area. The frequency gating margin is set according to the frequency resolution of short-time power spectrum estimation, the center frequency point extraction error of carrier frequency trajectory, and the residual frequency offset of the local oscillator. The corresponding burst envelope region is retained for the carrier frequency trajectory that passes both directional consistency comparison and amplitude fall area comparison, forming a Doppler drift gating observation frame.
[0034] It should be noted that the amplitude fall-off comparison uses the frequency range after the Doppler drift amplitude range is extended to both sides by the frequency gating margin as the comparison range, and the measured frequency drift amplitude falling into the comparison range is judged to meet the amplitude fall-off comparison.
[0035] S2.4: For each burst object in the Doppler drift-gated observation frame, check the Starlink ephemeris visible window, measured frequency drift direction, measured frequency drift amplitude, and burst duration. Burst objects that do not meet the Starlink ephemeris visible window requirement are removed by time, burst objects that do not meet the measured frequency drift direction requirement are removed by drift, burst objects that do not meet the measured frequency drift amplitude requirement are removed by amplitude, and burst objects that do not meet the burst duration continuity requirement are removed by continuity, thus forming a set of burst objects with consistent ephemeris. Furthermore, non-ephemeral consistency interference is eliminated one by one for burst objects in the Doppler drift-gated observation frames. First, the burst start time and burst end time of the burst object are compared with the Starlink ephemeral visible window. Burst objects that do not fall into the Starlink ephemeral visible window and whose overlapping time period is less than the minimum gating duration are eliminated by time. The measured frequency drift direction of the sudden object is compared with the Doppler drift direction of the candidate visible satellite, and sudden objects whose measured frequency drift direction is opposite to the Doppler drift direction are drifted out. The measured frequency drift amplitude of the burst object is compared with the Doppler drift amplitude range and frequency gating margin. Burst objects whose measured frequency drift amplitude does not fall within the Doppler drift amplitude range and frequency gating margin are eliminated. Then, the burst duration, burst interval, and carrier frequency trajectory endpoint difference between adjacent burst objects are checked according to the observation time. Burst objects whose burst interval exceeds the continuous transmission interval or whose carrier frequency trajectory endpoint difference exceeds the continuous frequency range are eliminated. Burst objects that pass the time check, drift check, amplitude check, and continuity check are retained to form an ephemeris consistent burst object set.
[0036] It should be noted that the continuous transmission interval is jointly set according to the time step of the continuous time-frequency analysis window, the average duration of the burst envelope region, and the number of adjacent continuous time-frequency analysis windows that are allowed to be missing.
[0037] S2.5: Merge the ephemeris consistent burst object set according to observation time, frequency band position and candidate visible satellites. Merge burst objects within the same transit time according to the adjacency of observation time, merge burst objects within the same frequency band according to the adjacency of frequency band position, and merge burst objects under the same candidate visible satellite according to the affiliation of candidate visible satellites to form terminal uplink candidate clusters; Furthermore, the set of burst objects with consistent ephemeris is merged according to observation time, frequency band location and candidate visible satellites. First, the candidate visible satellites are used as the first-level merging objects, and burst objects belonging to the same candidate visible satellite are classified into the same candidate visible satellite object group. For burst objects that belong to multiple candidate visible satellites, they are sorted according to the drift difference between the measured frequency drift amplitude and the Doppler drift amplitude range. The candidate visible satellite with the lowest drift difference is selected as the candidate visible satellite. Burst objects with the same drift difference are then sorted according to the time overlap ratio between the burst envelope region and the Starlink ephemeris visible window. The candidate visible satellite with the highest time overlap ratio is selected as the candidate visible satellite. Within each candidate visible satellite object group, burst objects within the same transit time period are merged according to the adjacent relationship of observation time, burst objects within the same frequency band are merged according to the adjacent relationship of frequency band location, and burst objects with frequency continuity are connected according to the difference of the endpoints of the carrier frequency trajectory. In each group of burst objects after merging, the burst envelope region, carrier frequency trajectory, measured frequency drift direction, measured frequency drift amplitude, burst duration, burst interval and candidate visible satellites are retained to form the terminal uplink candidate cluster.
[0038] It should be noted that the Starlink ephemeris visibility window is not simply a time window. It simultaneously includes elevation entry time, elevation exit time, azimuth sweep interval, and radial velocity variation interval. Ephemeris time gating is used to exclude bursts that are not within the transit time of candidate visible satellites, while Doppler drift gating is used to exclude bursts whose frequency drift direction and amplitude do not conform to the radial motion patterns of candidate visible satellites. For burst envelope regions crossing the boundaries of the Starlink ephemeris visibility window, overlapping time periods are used to prevent the overall deletion of boundary bursts. For bursts belonging to multiple candidate visible satellites, drift difference and time overlap ratio are used to sort the candidate visible satellites, preventing merging and confusion when multiple candidate visible satellites are simultaneously visible. The burst envelope regions and carrier frequency trajectories in the terminal uplink candidate clusters can continue to be used for the formation of subsequent continuous carrier frequency trajectories and ephemeris Doppler trajectories.
[0039] When the measured frequency drift amplitude falls within the range of the Doppler drift amplitude, the drift difference is taken as the difference between the measured frequency drift amplitude and the center value of the Doppler drift amplitude range; when the measured frequency drift amplitude does not fall within the range of the Doppler drift amplitude, the drift difference is taken as the difference between the measured frequency drift amplitude and the nearest boundary of the Doppler drift amplitude range.
[0040] S3: Continuously track the carrier frequency trajectory in the uplink candidate cluster of the terminal, calculate the ephemeris Doppler trajectory according to the candidate visible satellite orbit trajectory, perform Doppler inversion registration between the carrier frequency trajectory and the ephemeris Doppler trajectory, extract the uplink pre-compensation residual spectrum, and perform segmented Kalman smoothing and abrupt elimination on the uplink pre-compensation residual spectrum according to the burst envelope boundary to form the pre-compensation inversion trajectory; S3.1: Based on the burst objects in the uplink candidate cluster of the terminal, extract the carrier frequency trajectory corresponding to each burst object along the observation time, and continue the trajectory according to the frequency band proximity relationship and the consistent frequency drift direction relationship between adjacent burst objects to form a continuous carrier frequency trajectory; Furthermore, based on the burst objects in the terminal uplink candidate cluster, the terminal uplink candidate cluster is grouped according to the candidate visible satellites, and the burst objects are sorted according to the observation time under the same candidate visible satellite; For each burst, extract the burst envelope region, burst start time, burst end time, burst occupied frequency band, and carrier frequency trajectory. First, check the burst interval between adjacent bursts. Bursts whose burst intervals are within a continuous transmission interval are listed as continuable bursts. Then, check the frequency band proximity between continuable bursts. Bursts whose frequency band boundary interval between the burst occupied frequency band of the later burst and the burst occupied frequency band of the earlier burst do not exceed the frequency band proximity range are retained. Then, check the consistency of frequency drift direction between the retained bursts. Carrier frequency trajectories whose endpoint difference between the starting frequency point of the carrier frequency trajectory of the later burst and the ending frequency point of the carrier frequency trajectory of the earlier burst does not exceed the continuous frequency range and whose frequency drift direction is consistent are continuated to form a continuous carrier frequency trajectory.
[0041] S3.2: Map the observation time corresponding to the continuous carrier frequency trajectory to the candidate visible satellite trajectory, extract the radial velocity change of the candidate visible satellite relative to the monitoring point position during the observation time, and calculate the frequency offset change corresponding to the radial velocity change according to the target satellite communication frequency band to form the ephemeris Doppler trajectory. Furthermore, the observation time corresponding to the continuous carrier frequency trajectory is mapped to the candidate visible satellite trajectory. The spatial position and velocity of the candidate visible satellite are extracted from the candidate visible satellite trajectory according to the observation time in the continuous carrier frequency trajectory. The spatial position, velocity and monitoring point position of the candidate visible satellite are then converted to the same ground observation coordinates. The projection of the candidate visible satellite's velocity onto the line of sight is calculated along the line of sight from the monitoring point to the candidate visible satellite's spatial position. This forms the radial velocity change of the candidate visible satellite relative to the monitoring point's position during the observation time. The radial velocity change is then converted into a frequency offset change according to the target satellite's communication frequency band. The frequency offset change is then time-sampled according to the observation time of the continuous carrier frequency trajectory to form the ephemeris Doppler trajectory.
[0042] S3.3: Align the continuous carrier frequency trajectory and the ephemeris Doppler trajectory with the time axis according to the observation time, superimpose the ephemeris Doppler trajectory onto the continuous carrier frequency trajectory according to the reverse compensation relationship, and invert and subtract the frequency drift component caused by the motion of candidate visible satellites in the continuous carrier frequency trajectory to form the inverted registration frequency trajectory. Furthermore, the continuous carrier frequency trajectory and the ephemeris Doppler trajectory are aligned along the time axis according to the observation time. The frequency offset changes of the ephemeris Doppler trajectory that do not fall within the observation time points of the continuous carrier frequency trajectory are interpolated in time, so that the ephemeris Doppler trajectory and the continuous carrier frequency trajectory have one-to-one frequency point data at the same observation time point. Using the starting frequencies of the continuous carrier frequency trajectory and the ephemeris Doppler trajectory as alignment references, the fixed frequency offset in the continuous carrier frequency trajectory is subtracted. Then, the ephemeris Doppler trajectory is superimposed onto the continuous carrier frequency trajectory according to a reverse compensation relationship. The frequency drift component caused by the motion of candidate visible satellites in the continuous carrier frequency trajectory is inverted and subtracted to form the inverted registration frequency trajectory. Specific expressions include: ; in, This represents the residual frequency trajectory after inversion registration, i.e., the instantaneous value of the uplink pre-compensation residual; Indicates the observation time; Represents the frequency trajectory of a continuous carrier wave; The ephemeris Doppler frequency shift can be calculated from the radial velocity of the candidate visible satellites relative to the monitoring point. This represents a fixed frequency offset, specifically expressed as: ; in, Indicates the starting observation time of the continuous carrier frequency trajectory.
[0043] S3.4: Perform residual sampling on the inverted registration frequency trajectory along the observation time, map the residual sampling values to the time-frequency position of the corresponding burst object according to the burst envelope boundary, perform frequency distribution statistics on the residual sampling values within the same burst object, and form the uplink pre-compensation residual spectrum; Furthermore, residual sampling is performed on the inverted registration frequency trajectory along the observation time. The frequency offset corresponding to each observation time in the inverted registration frequency trajectory is used as the residual sampling value, and the residual sampling value is mapped to the time-frequency position of the corresponding burst object according to the burst envelope boundary. Within the same burst, the residual frequency grid is divided according to the frequency resolution of the short-time power spectrum estimation. The residual sampled values are then statistically analyzed according to the residual frequency grid, and the distribution position, distribution width, and energy concentration interval of the residual sampled values in the frequency dimension are recorded in combination with the energy weights corresponding to the residual sampled values, forming an uplink pre-compensated residual spectrum; the specific expressions include: ; in, The energy-weighted distribution function of residual sampled values within the same burst object in the frequency dimension; This represents the short-time power spectrum estimate corresponding to the location of the inverted registration frequency trajectory within the time-frequency analysis window; The function indicates that it takes the value 1 when the condition in parentheses is true, and 0 otherwise; the frequency grid represents the division of the residual frequency grid according to the frequency resolution of the short-time power spectrum estimation. This represents the set of observation times covered by the corresponding burst envelope region within the same burst object.
[0044] S3.5: Divide the uplink pre-compensated residual spectrum into multiple residual spectrum segments according to the burst envelope boundary. Within each residual spectrum segment, recursively calculate the residual state according to the observation time. Perform Kalman smoothing on the residual fluctuations between adjacent observation times to form a segmented smoothed residual spectrum. Furthermore, the uplink pre-compensated residual spectrum is divided into multiple residual spectrum segments according to the burst envelope boundary. Within each residual spectrum segment, a residual state sequence is established according to the observation time. The residual state sequence includes the residual center frequency and the residual drift rate. Using the residual center frequency and residual drift rate of the previous observation time, the residual state of the next observation time is recursively derived. Then, the residual sampled values in the uplink pre-compensated residual spectrum corresponding to the next observation time are used to correct the recursively derived residual state. This process of residual state recursion and correction is repeated within the residual spectrum segments. Kalman smoothing is applied to the residual fluctuations between adjacent observation times to form a segmented smoothed residual spectrum. Specific expressions include: State prediction equation: ; in, Indicates by the first The first state obtained by recursion at the observation time Predict the residual state at the observation time; Indicates the first The residual state after the observation time is updated; Represents the state transition matrix; Indicates the first Observation time and the The observation time interval between observation moments; State update equation: ; in, Indicates the first The residual state after the observation time is updated; Represents the Kalman gain matrix; Indicates the first The observation center frequency in the uplink pre-compensated residual spectrum at the observation time; This represents the observation matrix, used to extract the residual center frequency component from the residual state vector.
[0045] S3.6: Compare the change amplitude of adjacent residual states in the piecewise smooth residual spectrum along the observation time, remove residual states that are out of the continuity relationship of the burst envelope boundary and the continuity relationship of the carrier frequency trajectory, and reconnect the remaining residual states according to the observation time to form a pre-compensated inversion trajectory. Furthermore, the variation amplitudes of adjacent residual states in the piecewise smoothed residual spectrum are compared along the observation time. The residual center frequency jump variables and residual drift rate jump variables between adjacent residual states are checked against the continuity relationship of the burst envelope boundary, the continuity relationship of the carrier frequency trajectory, and the fluctuation range of the residual state after Kalman smoothing, respectively. Boundary abrupt changes are removed for residual states that deviate from the continuity relationship of the burst envelope boundary, trajectory abrupt changes are removed for residual states that deviate from the continuity relationship of the carrier frequency trajectory, and isolated points are removed for isolated residual states that exceed the fluctuation range of the residual state. The remaining residual states are then reconnected according to the observation time, burst envelope boundary, and candidate visible satellites to form a pre-compensated inversion trajectory.
[0046] It should be noted that Doppler inversion registration is not a simple comparison between continuous carrier frequency trajectories and ephemeris Doppler trajectories. Instead, it establishes a one-to-one correspondence between continuous carrier frequency trajectories and ephemeris Doppler trajectories at the same observation time point. After deducting the fixed frequency offset, it inverts and deducts the frequency drift component caused by the motion of candidate visible satellites according to the inverse compensation relationship. The uplink pre-compensation residual spectrum is not a single residual value, but a frequency distribution formed by statistical analysis of the residual frequency grid within the burst envelope boundary of the inverted registration frequency trajectory. The piecewise Kalman smoothing uses the residual center frequency and residual drift rate as the residual state. The abrupt elimination simultaneously verifies the continuity relationship of the burst envelope boundary and the continuity relationship of the carrier frequency trajectory. The pre-compensation inversion trajectory can continue to be used for the synchronous extraction of array phase difference sequences, array amplitude difference sequences, and angle of arrival estimation sequences.
[0047] The residual state fluctuation range is set jointly based on the predicted residual generated by piecewise Kalman smoothing, the distribution width of the uplink pre-compensated residual spectrum, and the frequency resolution of the short-time power spectrum estimation.
[0048] S4: Simultaneously extract the array phase difference sequence, array amplitude difference sequence, and angle of arrival estimation sequence from the pre-compensated inversion trajectory. Project the angle of arrival estimation sequence onto the geographic grid of the monitoring area. Then, perform angular frequency consistency closure verification according to the uplink pre-compensated residual spectrum, candidate visible satellite motion direction, burst envelope boundary continuity, and angle of arrival stability to form a Starlink terminal closure evidence chain. S4.1: Using the observation time and burst envelope boundary in the pre-compensated inversion trajectory as indexes, extract the array received data of the same burst object from the multi-channel intermediate frequency sampling sequence corresponding to the pre-compensated inversion trajectory, arrange the phase differences between each array received channel at the same observation time to form an array phase difference sequence, arrange the amplitude differences between each array received channel at the same observation time to form an array amplitude difference sequence, and perform angle of arrival estimation based on the array phase difference sequence and the array amplitude difference sequence to form an angle of arrival estimation sequence; Furthermore, using the observation time and burst envelope boundary in the pre-compensated inversion trajectory as indexes, array received data of the same burst object is extracted from the multi-channel intermediate frequency sampling sequence corresponding to the pre-compensated inversion trajectory. The array received data includes intermediate frequency sampling signals corresponding to each receiving channel at the same observation time, the same burst envelope boundary, and the same frequency band position. According to the channel identification order between the reference channel and other receiving channels, the phase of the intermediate frequency sampling signal of each receiving channel at the same observation time is extracted, and the phase difference of other receiving channels relative to the reference channel is arranged according to the observation time to form an array phase difference sequence. According to the channel identification order between the reference channel and other receiving channels, the amplitude of the intermediate frequency sampled signals of each receiving channel at the same observation time is extracted, and the amplitude differences of other receiving channels relative to the reference channel are arranged according to the observation time to form an array amplitude difference sequence. The array phase difference sequence is mapped to the array geometric position between the receiving channels, and the array amplitude difference sequence is used as the amplitude weight for the angle of arrival estimation. The angle of arrival direction of the same burst object at different observation times is estimated to form an angle of arrival estimation sequence. The specific expressions include: ; in, Indicates the first The estimated angle of arrival at the time of observation; The measured array phase difference vector is composed of the phase differences of each receiving channel relative to the reference channel at the same observation time. Indicates the angle of arrival The corresponding array phase difference theoretical model vector is determined by the array's geometric position and the signal direction; The measured array amplitude difference vector is composed of the amplitude differences of each receiving channel relative to the reference channel at the same observation time. Indicates the angle of arrival The corresponding array amplitude difference theoretical model vector is determined by the array pattern response; This represents the amplitude weighting coefficient, used to balance the contributions of the phase difference term and the amplitude difference term.
[0049] S4.2: Using the direction of the angle of arrival in the angle of arrival estimation sequence as the projection direction and the grid position in the geographic grid of the monitoring area as the projection object, the direction of the angle of arrival and the grid position in the geographic grid of the monitoring area are intersected angularly, and the observation time and burst envelope boundary corresponding to the angular intersection position are marked to the geographic grid of the monitoring area to form the angle of arrival grid projection map. Furthermore, taking the direction of the angle of arrival in the angle of arrival estimation sequence as the projection direction and the grid position in the geographic grid of the monitoring area as the projection object, the monitoring area is first divided into multiple grid positions according to the preset grid spacing, and the geographic coordinates and grid labels are marked for each grid position; Taking the location of the receiving channel as the projection starting point, each direction of the angle of arrival in the angle of arrival estimation sequence is extended to the geographic grid of the monitoring area, and the angular intersection calculation is performed on the grid positions through which the angle of arrival direction passes. The angular intersection positions of the same sudden object falling into adjacent grid positions within a continuous observation period are aggregated, and the observation time, sudden envelope boundary, frequency band position and candidate visible satellites corresponding to the angular intersection positions are labeled to the geographic grid of the monitoring area to form an angle of arrival grid projection map.
[0050] It should be noted that the preset grid spacing is set jointly based on the lateral position resolution corresponding to the angle resolution of the angle of arrival estimation sequence within the monitoring area, the positioning accuracy of the receiving channel, and the area of the monitoring area.
[0051] S4.3: Time-align the uplink pre-compensated residual spectrum corresponding to the angle of arrival grid projection map with the motion direction of the candidate visible satellite. Compare the residual change direction of the uplink pre-compensated residual spectrum with the Doppler change direction corresponding to the motion direction of the candidate visible satellite along the observation time. Write the observation time with matching directions into the frequency closure mark to form a frequency closure mark sequence. Furthermore, the uplink pre-compensation residual spectrum corresponding to the angle of arrival grid projection map is time-aligned with the motion direction of the candidate visible satellites. First, the uplink pre-compensation residual spectrum corresponding to the same burst object is retrieved according to the observation time in the angle of arrival grid projection map. Then, the Doppler change direction corresponding to the candidate visible satellites at the same observation time is extracted according to the motion direction of the candidate visible satellites. The direction of change of the residual center frequency in the uplink pre-compensated residual spectrum is calculated along the observation time, and the direction of change of the residual center frequency is checked against the direction of Doppler change. Observation times where the direction of change of the residual center frequency and the direction of Doppler change satisfy the reverse compensation relationship are written into frequency closure markers; observation times where the direction of change of the residual center frequency and the direction of Doppler change do not satisfy the reverse compensation relationship are not written into frequency closure markers. The frequency closure markers are arranged according to the observation time to form a frequency closure marker sequence. The specific expression for the determination condition of the frequency closure marker includes: ; in, This indicates the center frequency of the residual after piecewise Kalman smoothing; This represents a function that takes the value 1 when the independent variable is greater than 0, -1 when it is less than 0, and 0 when it is equal to 0. Indicates the ephemeris Doppler frequency; Indicates the first The frequency closure flag at the observation time has a value of 1 indicating that the frequency is closed and a value of 0 indicating that the frequency is not closed.
[0052] S4.4: Time-align the continuity of the burst envelope boundary corresponding to the frequency closure mark sequence with the angle of arrival grid projection map, check the continuity relationship of the burst envelope boundary and the angle fluctuation range of the angle of arrival estimation sequence along adjacent burst objects, and write spatial closure marks for burst objects with continuous continuity relationship and stable angle fluctuation range to form a spatial closure mark sequence. Furthermore, the continuity of the burst envelope boundary corresponding to the frequency closure marker sequence is time-aligned with the angle of arrival grid projection map. First, the burst envelope boundary continuity relationship between adjacent burst objects is extracted according to the observation time. The burst envelope boundary continuity relationship includes the burst termination time of the previous burst object, the burst start time of the next burst object, the burst interval between the previous and next burst objects, and the frequency band position difference between the previous and next burst objects. Then, the angular intersection position corresponding to the adjacent burst objects and the angular fluctuation range of the angle of arrival estimation sequence are extracted from the angle of arrival grid projection map. For burst objects whose burst interval is within a continuous transmission interval, whose frequency band position difference is within a continuous frequency range, whose angular intersection position is within an adjacent grid position, and whose angular fluctuation range is within a stable range of arrival angle, write spatial closure markers. Arrange the spatial closure markers according to the observation time to form a spatial closure marker sequence.
[0053] It should be noted that the stable range of the angle of arrival is set jointly based on the array geometric position error, the phase measurement error of the array phase difference sequence, and the angular resolution of the angle of arrival estimation sequence; The specific expressions for the determination conditions of the spatial closure marker include: ; in, Indicates the first The first sudden target and the first A spatial closure flag is set between burst objects, with a value of 1 indicating spatial closure and a value of 0 indicating spatial non-closure. Indicates the burst interval between two adjacent burst objects; Indicates the continuous transmission interval threshold; This represents the frequency band position difference between two adjacent burst objects; Indicates the threshold for a continuous frequency range; This represents the range of angular fluctuations in the estimated angle of arrival sequence between two adjacent burst objects; Indicates the threshold value for the stable range of the angle of arrival; Represents the logical AND operation.
[0054] S4.5: The frequency closure marker sequence and the spatial closure marker sequence are merged in the same position according to the observation time, the burst envelope boundary and the geographical grid of the monitoring area. The uplink pre-compensation residual spectrum, the motion direction of candidate visible satellites, the continuity of the burst envelope boundary and the stability of the angle of arrival in the same burst object are connected in a closed relationship to form a Starlink terminal closure evidence chain. Furthermore, the frequency closure marker sequence and the spatial closure marker sequence are merged in tandem according to observation time, burst envelope boundary, and monitoring area geographic grid. First, the correspondence between observation time, burst envelope boundary, uplink pre-compensation residual spectrum, candidate visible satellite motion direction, angle of arrival grid projection map, frequency closure marker sequence, and spatial closure marker sequence is established within the same burst object. Then, the frequency closure marker and spatial closure marker of the same burst object are concatenated in the order of observation time. The uplink pre-compensation residual spectrum, candidate visible satellite motion direction, burst envelope boundary continuity, and angle of arrival stability of the same burst object are concatenated with closure relationships to form a Starlink terminal closure evidence chain.
[0055] S5: Perform temporal aggregation of the Starlink terminal closed evidence chain in chronological order, perform finite state transition verification of adjacent Starlink terminal closed evidence chains, compile Starlink terminal detection evidence index, and output Starlink terminal wireless signal detection record. S5.1: Using the observation time in the Starlink terminal closure evidence chain as the main order and the candidate visible satellite, frequency band position, and monitoring area geographic grid as the secondary order, Starlink terminal closure evidence chains that belong to the same candidate visible satellite and have continuous observation time are sequentially arranged and adjacently labeled to form adjacent Starlink terminal closure evidence chain groups. Furthermore, the continuous observation interval is jointly set according to the time step of the continuous time-frequency analysis window, the average duration of the burst envelope region, and the number of observation windows that are allowed to be interrupted in the Starlink terminal closure evidence chain.
[0056] The Starlink terminal closure evidence chains are sorted by observation time as the primary order and candidate visible satellites, frequency band positions, and monitoring area geographic grids as secondary orders. First, the Starlink terminal closure evidence chains are grouped according to candidate visible satellites. Under the same candidate visible satellite, the Starlink terminal closure evidence chains are arranged in chronological order of observation time. Then, the observation time intervals, frequency band position differences, and monitoring area geographic grid position differences between adjacent Starlink terminal closure evidence chains are checked for adjacency. Starlink terminal closure evidence chains whose observation time intervals are within continuous observation intervals, whose frequency band position differences are within continuous frequency ranges, and whose monitoring area geographic grid position differences are within adjacent grid ranges are written into the same adjacency sequence, forming adjacent Starlink terminal closure evidence chain groups.
[0057] S5.2: According to the frequency closure marker sequence, spatial closure marker sequence, burst envelope boundary continuity relationship and uplink pre-compensation residual spectrum continuity relationship in the adjacent Starlink terminal closure evidence chain group, mark the capture state, stable transmission state, beam switching state and disappearance state for each Starlink terminal closure evidence chain, and check the transition order between adjacent states according to the observation time to form a finite state transition sequence; Furthermore, the state of each Starlink terminal closure evidence chain is marked according to the frequency closure marker sequence, spatial closure marker sequence, burst envelope boundary continuity relationship, and uplink pre-compensation residual spectrum continuity relationship in the adjacent Starlink terminal closure evidence chain groups; The Starlink terminal closure evidence chain that first appears a continuous frequency closure marker and begins to appear a spatial closure marker in an adjacent Starlink terminal closure evidence chain group is marked as the capture state. The Starlink terminal closure evidence chain that has both frequency closure markers and spatial closure markers for multiple consecutive observation times, has a continuous burst envelope boundary continuity relationship, and has a continuous uplink pre-compensation residual spectrum continuity relationship is marked as the stable launch state. The Starlink terminal closure evidence chain that maintains continuous frequency closure markers, where adjacent grids migrate in the monitoring area corresponding to the spatial closure markers, and where the burst envelope boundary continuity is restored before and after the migration is marked as a beam switching state; the Starlink terminal closure evidence chain that is interrupted in the continuous observation interval between frequency closure markers and spatial closure markers, and where the burst envelope boundary continuity is not restored in subsequent observation time is marked as a disappearing state; and the chronological relationship between the acquisition state, stable transmission state, beam switching state, and disappearing state is checked according to the observation time to form a finite state transition sequence.
[0058] S5.3: Map the finite state transition sequence to adjacent Starlink terminal closed evidence chain groups according to the observation time. For each Starlink terminal closed evidence chain, record the candidate visible satellites, frequency band positions, monitoring area geographic grid, frequency closure marker sequence, spatial closure marker sequence, and status annotation to form a Starlink terminal detection evidence index. Then, organize the detection time, detection frequency band, candidate visible satellites, terminal transmission status, detection confidence level, and alarm level according to the status annotation, frequency closure marker sequence, and spatial closure marker sequence in the Starlink terminal detection evidence index to form a Starlink terminal wireless signal detection record. Furthermore, the finite state transition sequences are mapped to adjacent Starlink terminal closure evidence chains according to observation time. For each Starlink terminal closure evidence chain, candidate visible satellites, frequency band positions, monitoring area geographic grids, frequency closure marker sequences, spatial closure marker sequences, and state annotations are recorded to form a Starlink terminal detection evidence index. Then, using the Starlink terminal detection evidence index as the retrieval entry point, the corresponding Starlink terminal closure evidence chains are extracted according to candidate visible satellites and observation time. The detection confidence is calculated based on the frequency closure marker sequence, spatial closure marker sequence, stable transmission state duration, and uplink pre-compensation residual spectrum continuity relationship, expressed as: ; in, To test the confidence level; The number of observation times that simultaneously possess both frequency closure and spatial closure markers; This represents the total number of observation times in the Starlink terminal closure evidence chain. To ensure the duration of a stable launch state; The duration of the Starlink terminal closure evidence chain; The residual continuation reference value is the sum of the frequency resolution of the short-time power spectrum estimation and the width of the uplink pre-compensated residual spectrum distribution. This represents the average residual difference between adjacent uplink pre-compensated residual spectra; The total number of observation times is obtained by statistically analyzing all observation times arranged by observation time within the same Starlink terminal closure evidence chain; the duration of stable launch state is obtained by accumulating the continuous observation times marked as stable launch state in the finite state transition sequence; the duration of the Starlink terminal closure evidence chain is obtained by the duration between the first and last observation times in the same Starlink terminal closure evidence chain; the residual continuation reference value is set according to the sum of the frequency resolution of the short-time power spectrum estimation and the distribution width of the uplink pre-compensation residual spectrum; the average residual difference between adjacent uplink pre-compensation residual spectra is obtained by averaging the residual center frequency differences of the uplink pre-compensation residual spectra corresponding to adjacent observation times in the same Starlink terminal closure evidence chain.
[0059] The detection confidence level is divided into low-confidence intervals, medium-confidence intervals, and high-confidence intervals according to the detection confidence level of historical confirmed samples, from low to high. The detection time, detection frequency band, candidate visible satellites, terminal transmission status, detection confidence level, and alarm level are combined to form a Starlink terminal wireless signal detection record.
[0060] It should be noted that adjacent Starlink terminal closed evidence chains are checked by adjacency verification through observation time intervals, frequency band position differences, and geographical grid position differences in the monitoring area, avoiding the mixing of evidence chains from different candidate visible satellites, different frequency band positions, and different geographical grids; finite state transition sequences are jointly formed by frequency closure marker sequences, spatial closure marker sequences, burst envelope boundary continuity relationships, and uplink pre-compensation residual spectrum continuity relationships, with acquisition state, stable transmission state, beam switching state, and disappearance state all corresponding to specific evidence sources; the Starlink terminal detection evidence index retains the mapping relationship between Starlink terminal closed evidence chains and Starlink terminal wireless signal detection records, facilitating subsequent verification of detection time, detection frequency band, candidate visible satellites, terminal transmission status, detection confidence, and alarm level.
[0061] See Figure 3 Under different signal-to-noise ratio (SNR) groupings, the detection F1 scores of the broadband time-frequency burst detection method (time-frequency detection), the ephemeris time-gated detection method (ephemeris gated), the ephemeris Doppler residual detection method (residual detection), and the method of this invention are compared, and the F1 improvement of the method of this invention compared to the ephemeris Doppler residual detection method is statistically analyzed. Figure 3 It can be seen that the method of the present invention maintains a high detection F1 under both low signal-to-noise ratio and medium-to-high signal-to-noise ratio conditions, indicating that star ephemeris gating, Doppler inversion registration and angular frequency consistency closure verification can improve the stability of Starlink terminal wireless signal detection.
[0062] See Figure 4 A standardized comparison was performed on detection F1, interference suppression rate, residual control capability, angle of arrival projection stability, angular frequency closure score, state transition accuracy, and detection confidence. Figure 4 It can be seen that the method of the present invention maintains a high level in multiple evaluation indicators, indicating that the reliability and evidence integrity of the Starlink terminal wireless signal detection record can be improved through continuous processing of pre-compensated inversion trajectory, Starlink terminal closed evidence chain and finite state transition verification.
[0063] Example 2: This embodiment provides a detection system for Starlink terminal wireless signals, including: The sampling observation module is used to collect multi-channel intermediate frequency sampling sequences of the target satellite communication frequency band within the monitoring area, perform synchronous sampling alignment, carrier frequency correction and array channel amplitude and phase calibration on the multi-channel intermediate frequency sampling sequences, and perform short-time power spectrum estimation, burst envelope segmentation and carrier frequency trajectory tracking through the time-frequency analysis window to form time-frequency burst observation frames. The ephemeris gating module is used to calculate the Starlink ephemeris visible window according to the monitoring point location, monitoring time, and low-Earth orbit satellite orbit information. It performs ephemeris time gating and Doppler drift gating on the time-frequency burst observation frames and the Starlink ephemeris visible window to eliminate non-ephemeris consistent interference and form a terminal uplink candidate cluster. The residual inversion module is used to continuously track the carrier frequency trajectory in the uplink candidate cluster of the terminal, calculate the ephemeris Doppler trajectory according to the orbital trajectory of the candidate visible satellite, perform Doppler inversion registration between the carrier frequency trajectory and the ephemeris Doppler trajectory, extract the uplink pre-compensated residual spectrum, and perform segmented Kalman smoothing and abrupt elimination on the uplink pre-compensated residual spectrum according to the burst envelope boundary to form the pre-compensated inversion trajectory; The angular frequency closure module is used to synchronously extract the array phase difference sequence, array amplitude difference sequence, and angle of arrival estimation sequence from the pre-compensated inversion trajectory. The angle of arrival estimation sequence is projected onto the geographic grid of the monitoring area. Then, the angular frequency consistency closure verification is performed according to the uplink pre-compensated residual spectrum, the motion direction of candidate visible satellites, the continuity of the burst envelope boundary, and the stability of the angle of arrival, forming a Starlink terminal closure evidence chain. The evidence index module is used to perform temporal aggregation of Starlink terminal closed evidence chains in chronological order, perform finite state transition verification of adjacent Starlink terminal closed evidence chains, compile Starlink terminal detection evidence index, and output Starlink terminal wireless signal detection records.
[0064] In summary, this invention continuously tracks the carrier frequency trajectory in the uplink candidate cluster of the terminal and performs Doppler inversion registration with the ephemeris Doppler trajectory to extract the uplink pre-compensated residual spectrum and form a pre-compensated inversion trajectory, thereby separating the terminal's uplink transmission characteristics from the motion factors of the low-Earth orbit satellite. Furthermore, by simultaneously extracting the array phase difference sequence, array amplitude difference sequence, and angle of arrival estimation sequence from the pre-compensated inversion trajectory and performing angular frequency consistency closure verification, a Starlink terminal closed evidence chain is formed. This chain is used for compiling the Starlink terminal detection evidence index and outputting the Starlink terminal wireless signal detection record, enabling the detection chain to have continuous tracking, spatial positioning, and status verification capabilities.
[0065] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of protection of the claims of the present invention.
Claims
1. A method for detecting a wireless signal of a satellite chain terminal, characterized in that, include: The system collects multi-channel intermediate frequency (IF) sampling sequences of the target satellite communication frequency band within the monitoring area. It performs synchronous sampling alignment, carrier frequency correction, and array channel amplitude and phase calibration on the multi-channel IF sampling sequences. It also performs short-time power spectrum estimation, burst envelope segmentation, and carrier frequency trajectory tracking through a time-frequency analysis window to form a time-frequency burst observation frame. The Starlink ephemeris visible window is calculated based on the location of the monitoring point, the monitoring time, and the low-Earth orbit information of the satellite. The time-frequency burst observation frames are subjected to ephemeris time gating and Doppler drift gating with the Starlink ephemeris visible window to eliminate non-ephemeris consistent interference and form a candidate cluster for terminal uplink. The carrier frequency trajectory in the uplink candidate cluster of the terminal is continuously tracked. The ephemeris Doppler trajectory is calculated according to the orbital trajectory of the candidate visible satellite. The carrier frequency trajectory and the ephemeris Doppler trajectory are Doppler inversion registration, the uplink pre-compensation residual spectrum is extracted, and the uplink pre-compensation residual spectrum is segmented Kalman smoothing and abrupt elimination according to the burst envelope boundary to form the pre-compensation inversion trajectory. The array phase difference sequence, array amplitude difference sequence, and angle of arrival estimation sequence are extracted synchronously from the pre-compensated inversion trajectory. The angle of arrival estimation sequence is projected onto the geographic grid of the monitoring area. Then, the angular frequency consistency closure check is performed according to the uplink pre-compensated residual spectrum, the motion direction of candidate visible satellites, the continuity of the burst envelope boundary, and the stability of the angle of arrival, forming a Starlink terminal closure evidence chain. The evidence chain of Starlink terminals is condensed in chronological order, the evidence chains of adjacent Starlink terminals are reviewed by finite state transitions, the Starlink terminal detection evidence index is compiled, and the Starlink terminal wireless signal detection record is output.
2. The method for detecting a wireless signal of a Starlink terminal according to claim 1, wherein, The formation of time-frequency burst observation frames is specifically as follows: Within the monitoring area, the target satellite communication frequency band is received synchronously through multiple channels. The radio frequency signals of each receiving channel are converted into intermediate frequency sampling signals. The intermediate frequency sampling signals are then uniformly organized according to the sampling time and channel identifier to form a multi-channel intermediate frequency sampling sequence. The multi-channel intermediate frequency sampling sequence is time-aligned between channels according to the sampling time. The carrier frequency of the time-aligned multi-channel intermediate frequency sampling sequence is corrected. The inter-channel amplitude difference correction and inter-channel phase difference correction are performed on the carrier frequency corrected multi-channel intermediate frequency sampling sequence with the reference channel as the reference, forming an array amplitude and phase calibration sequence. The array amplitude and phase calibration sequence is divided into continuous time-frequency analysis windows. Short-time power spectrum estimation is performed on the continuous time-frequency analysis windows. Energy mutation regions in the short-time power spectrum estimation are extracted along the observation time. The energy mutation regions are divided into burst envelope regions. The time continuous path of the center frequency point is tracked within the burst envelope region to form a time-frequency burst observation frame.
3. The method for detecting the wireless signal of the Starlink terminal according to claim 1, wherein, The process of applying ephemeris-time gating and Doppler drift gating to the time-frequency burst observation frames and the Starlink ephemeris visible window is as follows: The monitoring point location is converted into ground observation coordinates, and the low-orbit satellite orbit information is extrapolated according to the monitoring time to form a transit trajectory. The elevation angle entry time, elevation angle departure time, azimuth angle sweep interval and radial velocity change interval of candidate visible satellites are extracted along the transit trajectory to form the Starlink ephemeris visible window. The burst envelope region and carrier frequency trajectory are extracted from the time-frequency burst observation frame. The start and end times of the burst envelope region are overlapped with the elevation entry time and elevation exit time in the Starlink ephemeris visible window for screening. The corresponding carrier frequency trajectory is retained for the burst envelope region that falls into the Starlink ephemeris visible window, and the corresponding carrier frequency trajectory is removed for the burst envelope region that does not fall into the Starlink ephemeris visible window, thus forming the ephemeris time-gated observation frame. Based on the radial velocity variation range in the Starlink ephemeris visible window, the Doppler drift direction and Doppler drift amplitude range corresponding to the candidate visible satellites are calculated. The measured frequency drift direction and measured frequency drift amplitude are extracted along the carrier frequency trajectory in the ephemeris time-gated observation frame. The measured frequency drift direction is compared with the Doppler drift direction for directional consistency, and the measured frequency drift amplitude is compared with the Doppler drift amplitude range for amplitude fall area comparison, thus forming the Doppler drift-gated observation frame.
4. The method for detecting a wireless signal of a Starlink terminal according to claim 3, wherein The formation of the terminal uplink candidate cluster is as follows: For each burst object in the Doppler drift-gated observation frame, the Starlink ephemeris visibility window, measured frequency drift direction, measured frequency drift amplitude, and burst duration are checked. Burst objects that do not meet the Starlink ephemeris visibility window requirement are removed by time, burst objects that do not meet the measured frequency drift direction requirement are removed by drift, burst objects that do not meet the measured frequency drift amplitude requirement are removed by amplitude, and burst objects that do not meet the burst duration continuity requirement are removed by continuity, thus forming a set of burst objects with consistent ephemeris. The set of burst objects with consistent ephemeris is merged according to observation time, frequency band location and candidate visible satellites. Burst objects within the same transit time period are merged according to the adjacency of observation time, burst objects within the same frequency band are merged according to the adjacency of frequency band location, and burst objects under the same candidate visible satellite are merged according to the affiliation of candidate visible satellites, forming a terminal uplink candidate cluster.
5. The method for detecting Starlink terminal wireless signals as described in claim 1, characterized in that, The continuous tracking of carrier frequency trajectories in the candidate uplink clusters of the terminal, and the calculation of ephemeris Doppler trajectories based on the orbital trajectories of candidate visible satellites, are as follows: Based on the burst objects in the terminal uplink candidate cluster, the carrier frequency trajectory corresponding to each burst object is extracted along the observation time. The trajectory is then continued according to the frequency band proximity relationship and the consistent frequency drift direction relationship between adjacent burst objects to form a continuous carrier frequency trajectory. The observation time corresponding to the continuous carrier frequency trajectory is mapped to the orbital trajectory of the candidate visible satellite. The radial velocity change of the candidate visible satellite relative to the position of the monitoring point during the observation time is extracted, and the frequency offset change corresponding to the radial velocity change is calculated according to the communication frequency band of the target satellite to form the ephemeris Doppler trajectory.
6. The method for detecting Starlink terminal wireless signals as described in claim 5, characterized in that, The formation of the pre-compensated inversion trajectory is as follows: The continuous carrier frequency trajectory and the ephemeris Doppler trajectory are aligned along the time axis according to the observation time. The ephemeris Doppler trajectory is superimposed onto the continuous carrier frequency trajectory according to the reverse compensation relationship. The frequency drift component caused by the motion of candidate visible satellites in the continuous carrier frequency trajectory is inverted and subtracted to form the inverted registration frequency trajectory. Residual sampling is performed on the inverted registration frequency trajectory along the observation time. The residual sampled values are mapped to the time-frequency position of the corresponding burst object according to the burst envelope boundary. The frequency distribution statistics of the residual sampled values within the same burst object are performed to form the uplink pre-compensation residual spectrum. The uplink pre-compensated residual spectrum is divided into multiple residual spectrum segments according to the burst envelope boundary. Within each residual spectrum segment, the residual state is recursively deduced according to the observation time. Kalman smoothing is performed on the residual fluctuations between adjacent observation times to form a segmented smoothed residual spectrum. The variation amplitude of adjacent residual states in the piecewise smooth residual spectrum is compared along the observation time. Residual states that deviate from the continuity relationship of the burst envelope boundary and the continuity relationship of the carrier frequency trajectory are eliminated. The remaining residual states are reconnected according to the observation time to form a pre-compensated inversion trajectory.
7. The method for detecting Starlink terminal wireless signals as described in claim 1, characterized in that, The process of simultaneously extracting the array phase difference sequence, array amplitude difference sequence, and angle of arrival estimation sequence from the pre-compensated inversion trajectory, and projecting the angle of arrival estimation sequence onto the geographic grid of the monitoring area, is as follows: Using the observation time and burst envelope boundary in the pre-compensated inversion trajectory as indexes, array received data of the same burst object are extracted from the multi-channel intermediate frequency sampling sequence corresponding to the pre-compensated inversion trajectory. The phase differences between each array received channel at the same observation time are arranged to form an array phase difference sequence, and the amplitude differences between each array received channel at the same observation time are arranged to form an array amplitude difference sequence. The angle of arrival is estimated based on the array phase difference sequence and the array amplitude difference sequence to form an angle of arrival estimation sequence. Using the direction of the angle of arrival in the angle of arrival estimation sequence as the projection direction and the grid position in the geographic grid of the monitoring area as the projection object, the angle of arrival direction and the grid position in the geographic grid of the monitoring area are intersected angularly, and the observation time and burst envelope boundary corresponding to the angular intersection position are marked on the geographic grid of the monitoring area to form the angle of arrival grid projection map.
8. The method for detecting Starlink terminal wireless signals as described in claim 7, characterized in that, The formation of a closed chain of evidence for the Starlink terminal is as follows: The uplink pre-compensated residual spectrum corresponding to the angle of arrival grid projection map is time-aligned with the motion direction of the candidate visible satellite. The residual change direction of the uplink pre-compensated residual spectrum and the Doppler change direction corresponding to the motion direction of the candidate visible satellite are compared along the observation time. The observation times with matching directions are written into frequency closure markers to form a frequency closure marker sequence. The continuity of the burst envelope boundary corresponding to the frequency closure mark sequence is time-aligned with the angle of arrival grid projection map. The continuity relationship of the burst envelope boundary and the angle fluctuation range of the angle of arrival estimation sequence are checked along adjacent burst objects. Burst objects with continuous continuity relationship and stable angle fluctuation range are written into spatial closure marks to form a spatial closure mark sequence. The frequency closure marker sequence and the spatial closure marker sequence are merged in the same location according to the observation time, the burst envelope boundary and the geographical grid of the monitoring area. The uplink pre-compensation residual spectrum, the motion direction of candidate visible satellites, the continuity of the burst envelope boundary and the stability of the angle of arrival in the same burst object are connected in a closure relationship to form a Starlink terminal closure evidence chain.
9. The method for detecting Starlink terminal wireless signals as described in claim 1, characterized in that, The output Starlink terminal wireless signal detection record is as follows: Using the observation time in the Starlink terminal closure evidence chain as the primary order and the candidate visible satellite, frequency band location, and monitoring area geographic grid as the secondary order, Starlink terminal closure evidence chains that belong to the same candidate visible satellite and have continuous observation time are sequentially arranged and adjacently labeled to form adjacent Starlink terminal closure evidence chain groups. According to the frequency closure marker sequence, spatial closure marker sequence, burst envelope boundary continuity relationship, and uplink pre-compensation residual spectrum continuity relationship in the adjacent Starlink terminal closure evidence chain group, the acquisition state, stable transmission state, beam switching state, and disappearance state are marked for each Starlink terminal closure evidence chain. The transition order between adjacent states is checked according to the observation time to form a finite state transition sequence. The finite state transition sequence is mapped to adjacent Starlink terminal closure evidence chain groups according to the observation time. For each Starlink terminal closure evidence chain, candidate visible satellites, frequency band positions, monitoring area geographic grids, frequency closure marker sequences, spatial closure marker sequences, and status labels are recorded to form a Starlink terminal detection evidence index. The detection time, detection frequency band, candidate visible satellites, terminal transmission status, detection confidence level, and alarm level are organized according to the status labels, frequency closure marker sequences, and spatial closure marker sequences in the Starlink terminal detection evidence index to form a Starlink terminal wireless signal detection record.
10. A detection system for Starlink terminal wireless signals employing the detection method described in any one of claims 1 to 9, characterized in that, include: The sampling observation module is used to collect multi-channel intermediate frequency sampling sequences of the target satellite communication frequency band within the monitoring area, perform synchronous sampling alignment, carrier frequency correction and array channel amplitude and phase calibration on the multi-channel intermediate frequency sampling sequences, and perform short-time power spectrum estimation, burst envelope segmentation and carrier frequency trajectory tracking through the time-frequency analysis window to form time-frequency burst observation frames. The ephemeris gating module is used to calculate the Starlink ephemeris visible window according to the monitoring point location, monitoring time, and low-Earth orbit satellite orbit information. It performs ephemeris time gating and Doppler drift gating on the time-frequency burst observation frames and the Starlink ephemeris visible window to eliminate non-ephemeris consistent interference and form a terminal uplink candidate cluster. The residual inversion module is used to continuously track the carrier frequency trajectory in the uplink candidate cluster of the terminal, calculate the ephemeris Doppler trajectory according to the orbital trajectory of the candidate visible satellite, perform Doppler inversion registration between the carrier frequency trajectory and the ephemeris Doppler trajectory, extract the uplink pre-compensated residual spectrum, and perform segmented Kalman smoothing and abrupt elimination on the uplink pre-compensated residual spectrum according to the burst envelope boundary to form the pre-compensated inversion trajectory; The angular frequency closure module is used to synchronously extract the array phase difference sequence, array amplitude difference sequence, and angle of arrival estimation sequence from the pre-compensated inversion trajectory. The angle of arrival estimation sequence is projected onto the geographic grid of the monitoring area. Then, the angular frequency consistency closure verification is performed according to the uplink pre-compensated residual spectrum, the motion direction of candidate visible satellites, the continuity of the burst envelope boundary, and the stability of the angle of arrival, forming a Starlink terminal closure evidence chain. The evidence index module is used to perform temporal aggregation of Starlink terminal closed evidence chains in chronological order, perform finite state transition verification of adjacent Starlink terminal closed evidence chains, compile Starlink terminal detection evidence index, and output Starlink terminal wireless signal detection records.
