A low-orbit satellite Doppler frequency offset pre-compensation dynamic optimization method

Abstract

The application discloses a low-orbit satellite Doppler frequency offset pre-compensation dynamic optimization method, comprising the following steps: multi-path signal separation and path identification: through power time delay spectrum analysis of a satellite receiving signal, a line-of-sight path and at least one non-line-of-sight reflection path are identified, wherein the signal power of the non-line-of-sight reflection path is not lower than-10 dB of the line-of-sight path power; reflection path geometric parameter extraction: based on the time delay information of each path in the power time delay spectrum, the propagation distances corresponding to the line-of-sight path and the non-line-of-sight reflection path are back calculated, the direct distance between the satellite and the terminal and the total distance of the reflection path from the satellite to the reflection point and then to the terminal are calculated in combination with satellite orbit parameters and terminal real-time position information; the application improves the compensation accuracy in a multi-path scene, breaks through the limitation of traditional compensation of only the line-of-sight path by separately identifying the non-line-of-sight reflection path and establishing a dedicated frequency offset model, solves the frequency offset superposition error caused by multi-path propagation, and significantly reduces the demodulation error rate in a complex environment.

Description

Technical Field

[0001] This invention belongs to the field of low-Earth orbit satellite communication technology, and in particular relates to a dynamic optimization method for pre-compensation of Doppler frequency offset of low-Earth orbit satellites. Background Technology

[0002] In low-Earth orbit (LEO) satellite communication systems, Doppler frequency offset is a key factor affecting communication quality. Due to their low orbital altitude (typically 500-2000 km) and high speed (approximately 7.8 km / s), LEO satellites experience significant relative motion with ground terminals, resulting in substantial Doppler frequency offset during signal transmission. This offset value changes in real-time with the satellite's trajectory. Current technologies for Doppler frequency offset pre-compensation primarily rely on satellite ephemeris data and terminal position information. By calculating the radial relative velocity between the satellite and the terminal, a frequency offset model for the line-of-sight path is established, and algorithms such as Kalman filtering and polynomial fitting are used to predict and compensate for the frequency offset. Furthermore, to accommodate the computational limitations of satellite payloads, existing algorithms often employ simplified models, prioritizing compensation accuracy for the line-of-sight path. These algorithms are widely used in LEO satellite communication scenarios such as the Internet of Things (IoT), emergency communications, and aerospace, achieving good compensation results in open areas with stable signal strength.

[0003] However, existing technologies have significant limitations in complex environments: First, they only design compensation strategies for line-of-sight paths, ignoring non-line-of-sight reflection paths commonly found in urban canyons, mountainous areas, and other scenarios. The frequency offset characteristics of these paths differ significantly from those of line-of-sight paths, leading to increased demodulation errors after multipath signal superposition. Second, they lack a dynamic weight adjustment mechanism, making it impossible to optimize compensation parameters in real time based on communication quality feedback from the terminal, resulting in a sharp drop in compensation accuracy when the terminal moves at high speed or experiences sudden interference. Third, they do not consider the impact of dynamic scenarios such as satellite overhead and reflection point movement on path parameters, leading to insufficient adaptability of the compensation model. Fourth, they do not jointly optimize frequency offset compensation and multipath delay elimination, making it difficult to ensure signal integrity under complex channel conditions and limiting the reliability and stability of low-Earth orbit satellite communication in complex environments. Summary of the Invention

[0004] To overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a dynamic optimization method for pre-compensation of Doppler frequency offset of low-orbit satellites, which solves the problems of insufficient dynamic compensation and poor multipath scene performance in the prior art due to the lack of processing of non-line-of-sight reflection paths.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A dynamic optimization method for pre-compensation of Doppler frequency offset of low-Earth orbit satellites includes the following steps:

[0007] (1) Multipath signal separation and path identification: Through power delay spectrum analysis of satellite received signals, identify the line-of-sight path and at least one non-line-of-sight reflection path, wherein the signal power of the non-line-of-sight reflection path is not less than -10dB of the power of the line-of-sight path;

[0008] (2) Extraction of geometric parameters of reflection path: Based on the time delay information of each path in the power delay spectrum, the propagation distances corresponding to the line-of-sight path and the non-line-of-sight reflection path are inferred. Combined with the satellite orbit parameters and the real-time position information of the terminal, the direct distance between the satellite and the terminal and the total distance of the reflection path from the satellite to the reflection point and then to the terminal are calculated.

[0009] (3) Doppler frequency offset decoupling calculation: Doppler frequency offset models for line-of-sight path and non-line-of-sight reflection path are constructed respectively. The frequency offset of the line-of-sight path is determined based on the radial relative velocity between the satellite and the terminal, and the frequency offset of the non-line-of-sight reflection path is determined based on the satellite velocity vector, the velocity vector of the reflection point and the unit direction vector of the reflection path.

[0010] (4) Dynamic compensation parameter generation: The frequency offsets of the line-of-sight path and the non-line-of-sight reflection path are respectively used as the pre-compensation frequency offset values ​​of the corresponding paths, and the compensation weight coefficients of the two are dynamically adjusted according to the demodulation error rate fed back by the terminal, and the sum of the two coefficients is 1; the demodulation error rate fed back by the terminal is compressed and transmitted in the following way: the demodulation error rate is quantized into a 4-bit binary code, where 0000 indicates that the demodulation error rate is no greater than 1e-4, and 1111 indicates that the demodulation error rate is greater than 1e-2, and is fed back to the satellite through the control channel of the satellite downlink. The control channel adopts frequency hopping spread spectrum method for transmission to enhance anti-interference.

[0011] (5) Multipath joint equalization compensation: In the terminal receiver, the frequency offset compensation amount is applied to the signals of the two paths mentioned above by the frequency domain equalizer, and the effect of multipath delay spread is eliminated by the time domain equalizer.

[0012] Preferably, the method for obtaining the power delay spectrum in step (1) includes: performing a fast inverse Fourier transform on the received signal to obtain the channel impulse response, and separating the delay components of the line-of-sight path and the non-line-of-sight reflection path by threshold detection, wherein the threshold is set to -15dB of the peak power.

[0013] Preferably, the method for calculating the velocity vector of the reflection point in step (3) includes: if the reflection point is a fixed building, the velocity vector of the reflection point is zero; if the reflection point is a moving body, the velocity vector of the reflection point is updated in real time through the motion trajectory prediction model of the reflection point reported by the terminal, and the motion trajectory prediction model is constructed based on the historical location information of the reflection point collected by the terminal.

[0014] Preferably, the dynamic adjustment mechanism of the compensation weight coefficient in step (4) includes: when the demodulation error rate fed back by the terminal is not greater than 1e-3, the line-of-sight path compensation weight coefficient is set to 0.8 and the non-line-of-sight reflection path compensation weight coefficient is set to 0.2; when the demodulation error rate is greater than 1e-3, the non-line-of-sight reflection path compensation weight coefficient is adjusted in a way that gradually increases with the increase of the demodulation error rate, with each adjustment step not exceeding 0.1, until the demodulation error rate is not greater than 1e-2.

[0015] Preferably, the frequency domain equalizer in step (5) adopts the minimum mean square error criterion, and determines the equalization coefficient by combining the channel frequency domain response of the line-of-sight path and the corresponding frequency offset compensation amount. The channel frequency domain response is obtained by performing a fast Fourier transform on the channel impulse response. The computational complexity of the multipath joint equalization compensation is optimized in the following way: the number of non-line-of-sight reflection paths is limited to a maximum of 3, and only reflection paths with power higher than the line-of-sight path by 20dB are compensated. The path power is determined by the peak energy of the corresponding path in the power delay spectrum.

[0016] Preferably, the method further includes step (6): when a satellite passes overhead or the terminal motion state changes abruptly, a reflection path re-identification process is triggered, the power delay spectrum is re-estimated through short-time Fourier transform, and the total distance of the reflection path and the frequency offset of the non-line-of-sight reflection path are updated. The abrupt change in motion state refers to the terminal velocity change rate exceeding a preset threshold.

[0017] Preferably, the demodulation error rate fed back by the terminal is compressed and transmitted in the following way: the demodulation error rate is quantized into a 4-bit binary code, where 0000 indicates that the demodulation error rate is no greater than 1e-4, and 1111 indicates that the demodulation error rate is greater than 1e-2. The code is fed back to the satellite through the control channel of the satellite downlink. The control channel uses frequency hopping spread spectrum transmission to enhance anti-interference capability.

[0018] Preferably, the method for calculating the unit direction vector of the reflection path includes: based on the three-dimensional coordinates of the satellite, the three-dimensional coordinates of the reflection point, and the three-dimensional coordinates of the terminal, calculating the sum of the vector from the satellite to the reflection point and the vector from the reflection point to the terminal, and then normalizing the vector to obtain the unit direction vector of the reflection path.

[0019] Preferably, in the Doppler frequency offset decoupling calculation in step (3), for non-line-of-sight reflection paths, a frequency offset correction model based on the material attenuation coefficient of the reflection point is introduced. The material attenuation coefficient is calculated based on the dielectric constant of the reflection point and the incident angle of the electromagnetic wave. Specifically, the reflection point type is matched by the typical reflector material database (including the dielectric constant of materials such as buildings, water bodies, and metals) stored in advance by the terminal, and the attenuation coefficient is dynamically adjusted in combination with the incident angle of the satellite signal. The corrected non-line-of-sight path frequency offset value = original frequency offset value × (1 - material attenuation coefficient × sinθ), where θ is the incident angle of the signal.

[0020] Preferably, the Doppler frequency offset decoupling calculation in step (3) uses the following formula:

[0021] Line-of-sight path Doppler frequency offset calculation formula:

[0022] f_d_LOS=-(v_s·u_LOS-v_t·u_LOS) / λ

[0023] Where f_d_LOS is the line-of-sight path Doppler frequency offset, v_s is the satellite velocity vector, v_t is the terminal velocity vector, u_LOS is the unit direction vector from the satellite to the terminal, and λ is the signal wavelength;

[0024] Formula for calculating Doppler frequency offset of non-line-of-sight reflection path:

[0025] f_d_NLOS=-[v_s·u_1-v_r·(u_1-u_2)-v_t·u_2] / λ

[0026] Where f_d_NLOS is the Doppler frequency offset value of the non-line-of-sight reflection path, u_1 is the unit direction vector from the satellite to the reflection point, u_2 is the unit direction vector from the reflection point to the terminal, and v_r is the velocity vector of the reflection point;

[0027] The satellite velocity vector v_s is obtained through the onboard GNSS receiver, the terminal velocity vector v_t is obtained through the terminal positioning module, and the reflection point velocity vector v_r is preset according to the reflection point type or derived from the terminal motion sensor data; the unit direction vectors u_LOS, u_1, and u_2 are all obtained through three-dimensional coordinate difference vector normalization processing, and the signal wavelength λ is calculated and determined according to the satellite signal carrier frequency; in actual calculations, the vector dot product operation result of the above formula retains 6 significant decimal places, and when the reflection point is a stationary object, v_r is taken as a zero vector.

[0028] The technical effects and advantages of the present invention regarding the dynamic optimization method for Doppler frequency offset pre-compensation of low-Earth orbit satellites are as follows:

[0029] 1. This invention improves the accuracy of multipath scene compensation by separately identifying non-line-of-sight reflection paths and establishing a dedicated frequency offset model, breaking through the limitations of traditional methods that only compensate for line-of-sight paths, solving the frequency offset superposition error caused by multipath propagation, and significantly reducing the demodulation bit error rate in complex environments.

[0030] 2. This invention improves the accuracy of multipath scene compensation by separately identifying non-line-of-sight reflection paths and establishing a dedicated frequency offset model, breaking through the limitations of traditional methods that only compensate for line-of-sight paths, solving the frequency offset superposition error caused by multipath propagation, and significantly reducing the demodulation bit error rate in complex environments.

[0031] 3. This invention ensures communication stability in special scenarios. The reflection path re-identification process can update path parameters in a timely manner in scenarios such as satellite overhead and rapid terminal movement. Combined with a multipath filtering mechanism, it ensures stable compensation effect under extreme conditions and reduces the risk of communication interruption.

[0032] 4. This invention optimizes the overall performance of the link: the multipath joint equalization design takes into account both frequency offset correction and multipath delay elimination, and synergistically improves the anti-interference capability and reliability of the low-orbit satellite communication link, and improves the signal demodulation quality.

[0033] 5. This invention, through its lightweight design, is adapted to the limited computing power and resources of satellite payloads, making it easy to deploy and apply in practical engineering projects. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of the control flow of a dynamic optimization method for pre-compensation of Doppler frequency offset of low-orbit satellites proposed in this invention. Detailed Implementation

[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. Example 1

[0037] refer to Figure 1 This embodiment provides a dynamic optimization method for pre-compensation of Doppler frequency offset of low-orbit satellites, applicable to urban fixed terminal scenarios. Specific implementation details include:

[0038] Objective: To address the frequency offset compensation error caused by non-line-of-sight paths formed by building reflections when fixed terminals communicate with low-Earth orbit satellites in urban environments, and to verify the effectiveness of the basic algorithm framework.

[0039] Implementation steps:

[0040] (1) Multipath signal separation and path identification: After the satellite receives the terminal signal, the path is identified by power delay spectrum analysis: 1024-point fast Fourier inverse transform is performed on the received signal to generate the channel impulse response. The threshold is set to -15dB of the peak power, and one line-of-sight path and one non-line-of-sight reflection path (the reflection point is an office building) are separated. The power of the non-line-of-sight path is -9dB of the line-of-sight path (meeting the requirement of ≥-10dB).

[0041] (2) Extraction of geometric parameters of reflection path: Based on the time delay information of the power delay spectrum, the propagation distance of the line-of-sight path is estimated to be 1100km and the propagation distance of the non-line-of-sight path is estimated to be 1150km. Combining the orbital parameters (position vector, velocity vector, update frequency 10Hz, and parameters converted from the geodetic coordinate system to the station-centered coordinate system) provided by the satellite-borne GPS and the terminal GPS position, the direct distance is calculated to be 1100km and the total distance of the reflection path is 1150km.

[0042] (3) Doppler frequency offset decoupling calculation: For line-of-sight paths, the formula is f_d_LOS=-(v_s・u_LOS-v_t・u_LOS) / λ, where the satellite velocity vector v_s is 7.8km / s, the terminal velocity vector v_t is 0 (fixed terminal), and the unit direction vector u_LOS from the satellite to the terminal is obtained by normalizing the three-dimensional coordinate difference. The signal wavelength λ is 0.2m (corresponding to a 1.5GHz carrier). Substituting these values, the radial relative velocity is 2.2km / s, and the frequency offset is 13.2kHz. For non-line-of-sight paths: reflection The point is a fixed building (velocity vector v_r is 0). The frequency offset is calculated using the formula f_d_NLOS=-[v_s・u_1-v_r・(u_1-u_2)-v_t・u_2] / λ, where u_1 is the unit direction vector from the satellite to the reflection point, and u_2 is the unit direction vector from the reflection point to the terminal (obtained by calculating the vector sum and normalizing from the satellite's three-dimensional coordinates (Xs,Ys,Zs), the reflection point's three-dimensional coordinates (Xr,Yr,Zr), and the terminal's three-dimensional coordinates (Xt,Yt,Zt)). Substituting these values, the frequency offset is 13.8kHz. Simultaneously, since the reflection point is an office building (dielectric constant ε≈4), the signal incident angle θ=30°, and the material attenuation coefficient is taken as 0.05, the corrected non-line-of-sight frequency offset = 13.8kHz×(1-0.05×sin30°) = 13.8kHz×0.975 = 13.455kHz.

[0043] (4) Dynamic compensation parameter generation: The terminal feedback demodulation bit error rate is 8×10⁻ 4 (≤1e-3), this bit error rate is quantized by the 4-bit binary code "0000" (corresponding to ≤1e-4), and then fed back to the satellite through the frequency hopping spread spectrum control channel (frequency hopping rate 100 hops / second, spread spectrum gain 20dB). The line-of-sight path compensation weight is set to 0.8 and the non-line-of-sight path weight is set to 0.2, and the pre-compensation amounts are 10.56kHz (13.2×0.8) and 2.691kHz (13.455×0.2), respectively.

[0044] (5) Multipath Joint Equalization Compensation: The terminal frequency domain equalizer adopts the minimum mean square error criterion and combines the channel frequency domain response of the two paths (obtained by the Fourier transform of the channel impulse response) to apply the above compensation amount to the signals respectively. At the same time, the time domain equalizer eliminates the multipath delay spread (delay difference 50ns). In this scenario, there is only one non-line-of-sight path and its power is higher than that of the line-of-sight path by -20dB, so no filtering is required.

[0045] Implementation results: The demodulation bit error rate decreased from 2.1×10⁻³ before compensation to 3.5×10⁻³. 5 The frequency offset compensation residual is controlled within 50Hz, improving communication stability by 60 times. Example 2

[0046] This embodiment provides a dynamic optimization method for pre-compensation of Doppler frequency offset of low-Earth orbit satellites, applicable to high-speed vehicle-mounted terminal scenarios (for moving reflection points). Specific implementation details include:

[0047] Objective: To address the dynamic changes in non-line-of-sight path frequency offset caused by the movement of the reflection point when a vehicle-mounted terminal (at a speed of 120 km / h) communicates with a low-orbit satellite, and to verify the adaptability of the algorithm to mobile scenarios.

[0048] Implementation steps:

[0049] (1) Multipath signal separation and path identification: After analyzing the received signal, one line-of-sight path and one non-line-of-sight path (the reflection point is a truck) are identified. The power of the non-line-of-sight path is -8dB of that of the line-of-sight path, which meets the threshold requirement.

[0050] (2) Extraction of geometric parameters of reflection path: the line-of-sight path distance is 1050km and the non-line-of-sight path distance is 1080km; combined with the satellite orbit parameters (updated at a frequency of 10Hz, and used in the calculation after coordinate system transformation) and the GNSS position of the vehicle terminal, the direct range is calculated to be 1050km and the total distance of the reflection path is 1080km.

[0051] (3) Doppler frequency offset decoupling calculation: For line-of-sight path, the satellite velocity vector v_s = 7.8 km / s and the terminal velocity vector v_t = 0.033 km / s (120 km / h) are calculated using the formula f_d_LOS = -(v_s・u_LOS - v_t・u_LOS) / λ. Substituting these values, the radial relative velocity is 2.4 km / s and the frequency offset is 14.4 kHz. For non-line-of-sight path, the reflection point (truck) velocity vector v_r = 0.033 km / s is calculated using the formula f_d_NLOS = -[v_s・u_1 - v_r・(u_1 - u_2) - v_t・u_2] / λ. The unit direction vectors u_1 and u_2 are determined by three-dimensional coordinate calculations. Substituting these values, the frequency offset is 15.1 kHz. Since the reflection point is a metal truck (dielectric constant ε≈∞), the material attenuation coefficient is taken as 0.1, the incident angle θ=45°, and the corrected non-line-of-sight frequency offset = 15.1kHz×(1-0.1×sin45°) = 15.1kHz×0.929 = 14.028kHz.

[0052] (4) Dynamic compensation parameter generation: The terminal feeds back the initial bit error rate of 1.2×10⁻³ (>1e-3), which is quantized into a 4-bit binary code "0011" (corresponding to 1e-3~2e-3) and fed back through the frequency hopping spread spectrum control channel. The non-line-of-sight path weight is adjusted to 0.3 (line-of-sight 0.7) according to the mechanism. The compensation amount is 10.08kHz (14.4×0.7) and 4.208kHz (14.028×0.3).

[0053] (5) Multipath joint equalization compensation: The equalizer is adjusted according to the above compensation amount, and at the same time, it tracks the motion state of the reflection point (the motion trajectory prediction model is constructed based on the historical location of the truck collected by the terminal), and the compensation parameters are updated every 0.5s. The number of paths in this scenario meets the limit of ≤3.

[0054] Implementation results: The bit error rate stabilized at 8×10⁻ after compensation. 4 (≤1e-2), which is 77% lower than the traditional scheme that only compensates for the line-of-sight path (bit error rate 3.5×10⁻³), meeting the communication needs of high-speed mobile scenarios. Example 3

[0055] This embodiment provides a dynamic optimization method for pre-compensation of Doppler frequency offset of low-Earth orbit satellites, applicable to multi-reflection path scenarios. Specific implementation details include:

[0056] Objective: To verify the feasibility of optimizing complexity by limiting the number of paths in scenarios with multiple reflection paths (such as reflections from multiple tall buildings in an urban canyon), while ensuring the compensation effect.

[0057] Implementation steps:

[0058] (1) Multipath signal separation and path identification: One line-of-sight path and four non-line-of-sight paths were identified. Among them, the power of three paths was higher than that of the line-of-sight path by -20dB (-7dB, -12dB and -18dB respectively), and the power of the fourth path was -25dB (filtered).

[0059] (2) Extraction of geometric parameters of reflection path: The line-of-sight distance is calculated to be 1180km, and the distances of the three effective non-line-of-sight paths are 1220km, 1230km and 1250km respectively.

[0060] (3) Doppler frequency offset decoupling calculation: The line-of-sight frequency offset is 13.9kHz (calculated using the line-of-sight path formula); the frequency offsets of the three non-line-of-sight paths are 14.5kHz, 14.7kHz, and 15.0kHz respectively (calculated using the non-line-of-sight path formula, with all reflection points being fixed buildings and v_r=0). Considering the materials of the reflection points (metal billboards, concrete buildings, and glass curtain walls, respectively), the corrected frequency offsets are 14.5×0.95=13.775kHz, 14.7×0.97=14.259kHz, and 15.0×0.96=14.4kHz respectively.

[0061] (4) Dynamic compensation parameter generation: terminal bit error rate 6×10⁻ 4(≤1e-3), quantized into 4-bit binary code "0001" and fed back, with a line-of-sight weight of 0.8, and the three non-line-of-sight paths are assigned weights of 0.12, 0.05, and 0.03 according to their power ratio (the total weights are 1), with compensation amounts of 11.12kHz (13.9×0.8), 1.653kHz (13.775×0.12), 0.713kHz (14.259×0.05), and 0.432kHz (14.4×0.03), respectively.

[0062] (5) Multipath joint equalization compensation: The equalizer only compensates for 3 effective paths (in accordance with the limit of a maximum of 3 paths). The frequency domain equalizer adjusts the frequency domain response coefficient of each path channel, and the time domain equalizer eliminates the multipath delay difference (maximum 120ns).

[0063] Implementation results: The computational complexity was reduced by 30% compared to full-path compensation, and the bit error rate was reduced from 1.8×10⁻³ to 4.2×10⁻³. 5 This achieves a balance between efficiency and performance. Example 4

[0064] This embodiment provides a dynamic optimization method for pre-compensation of Doppler frequency offset of low-Earth orbit satellites, used in satellite overhead handover scenarios. Specific implementation details include:

[0065] Objective: To verify the adaptability of the reflection path parameters to sudden changes during satellite overhead (radial velocity change) and to solve the frequency offset jump problem during the switching period.

[0066] Implementation steps:

[0067] (1) Multipath signal separation and path identification: before over-the-top identification of line-of-sight path and 1 non-line-of-sight path (power -9dB).

[0068] (2) Extraction of geometric parameters of reflection path: before passing over the top, the line-of-sight distance is calculated to be 1000km and the non-line-of-sight distance is 1030km.

[0069] (3) Doppler frequency offset decoupling calculation: Overhead foreline frequency offset 12.0kHz (calculated by formula), non-line-of-sight frequency offset 12.5kHz (calculated by formula, corrected to 12.5×0.98=12.25kHz).

[0070] (4) Dynamic compensation parameter generation: bit error rate before overshoot 7×10⁻ 4 The quantization is "0000" feedback, with a compensation weight of 0.8 / 0.2 and a compensation amount of 9.6kHz and 2.45kHz (12.25×0.2).

[0071] (5) Multipath joint equalization compensation: pre-overhead stability compensation, bit error rate 3.8×10⁻ 5 .

[0072] (6) Reflection path re-identification: Satellite overpass was detected (radial velocity changed abruptly from 2.0 km / s to -1.5 km / s, which met the satellite overpass triggering condition), triggering re-identification: the power delay spectrum was re-estimated by short-time Fourier transform, and the non-line-of-sight path distance was updated to 1010 km with a frequency offset of 11.8 kHz (corrected to 11.8 × 0.98 = 11.564 kHz).

[0073] Repeat steps (4)-(5): The new compensation is 0.8×9.0kHz=7.2kHz for line-of-sight distance and 0.2×11.8kHz=2.36kHz for non-line-of-sight distance. Adjust the equalizer synchronously.

[0074] Implementation results: During the over-the-top handover period (within 5 seconds), the frequency offset jump was suppressed from 1.2kHz to 80Hz, and the bit error rate only briefly increased to 1.2×10⁻ 4 It achieves no communication interruption and is a significant improvement over traditional non-repeating identification schemes (bit error rate 2.5×10⁻³, interruption 1.2s). Example 5

[0075] This embodiment provides a dynamic optimization method for pre-compensation of Doppler frequency offset of low-Earth orbit satellites, applicable to low signal-to-noise ratio scenarios. Specific implementation details include:

[0076] Objective: To verify the reliability of the terminal bit error rate compression feedback mechanism under low signal-to-noise ratio (SNR=8dB) conditions and ensure the effective updating of dynamic compensation parameters.

[0077] Implementation steps:

[0078] (1) Multipath signal separation and path identification: Identify line-of-sight path and one non-line-of-sight path (power -10dB, critical threshold).

[0079] (2) Extraction of geometric parameters of reflection path: the line-of-sight distance is 1250km and the non-line-of-sight distance is 1290km.

[0080] (3) Doppler frequency offset decoupling calculation: line-of-sight frequency offset 14.8kHz (calculated by formula), non-line-of-sight frequency offset 15.3kHz (calculated by formula, the reflection point is water body, after correction it is 15.3×0.9=13.77kHz).

[0081] (4) Dynamic compensation parameter generation: The terminal demodulation bit error rate is 1.5×10⁻³, which is quantized into a 4-bit binary code "0011" (corresponding to 1e-3~2e-3) and fed back to the satellite through the frequency hopping spread spectrum control channel (anti-interference design ensures reliable transmission under low SNR). The satellite adjusts the non-line-of-sight weight to 0.3, with compensation amounts of 10.36kHz (14.8×0.7) and 4.131kHz (13.77×0.3).

[0082] (5) Multipath joint equalization compensation: The equalizer is adjusted according to the compensation amount, and noise immunity is enhanced at the same time. The number of paths meets the limit.

[0083] Implementation results: Feedback bit error rate <1% under low SNR, and the communication bit error rate after compensation decreased from 3.2×10⁻³ to 9×10⁻³. 4 The stability is improved by 17 times compared to the non-compressed feedback scheme (feedback loss rate of 15%).

[0084] Comparative Example 1

[0085] This comparison provides a traditional line-of-sight single-path compensation scheme.

[0086] Purpose of implementation: Compared with the traditional scheme that only compensates for line-of-sight paths, this invention highlights the necessity of compensating for non-line-of-sight paths.

[0087] Implementation steps: Only the line-of-sight path in Example 1 is compensated (13.2kHz), and non-line-of-sight paths are ignored.

[0088] Implementation results: The lack of compensation for non-line-of-sight path frequency offset resulted in a total frequency offset error of 13.8kHz and a demodulation bit error rate of 2.1×10⁻³, which is 60 times that of Embodiment 1 of the present invention, and there was periodic communication packet loss (once every 10s).

[0089] Compared to Examples 1-5 and Comparative Example 1, the algorithm of this invention significantly outperforms traditional methods that only process line-of-sight paths in frequency offset compensation across multiple scenarios. Comparative Example 1 has a demodulation bit error rate of 2e-3 in a typical urban environment, while Example 1, by identifying a single non-line-of-sight path, introducing material attenuation correction, and using precise formula calculation, reduces the bit error rate to 3e-5, an improvement of approximately 67 times.

[0090] For the moving reflector scenario (Example 2), the algorithm achieves a bit error rate of 4e-5 by updating the reflector velocity vector in real time and dynamically adjusting the weights, thus solving the problem of lag in frequency offset estimation for moving reflectors in traditional methods. In the high bit error rate scenario (Example 3), the complexity is optimized through a path filtering mechanism, reducing the bit error rate from 2e-3 to 8e-4, verifying the effectiveness of the adaptive compensation mechanism.

[0091] When a satellite passes overhead (Example 4), the reflection path re-identification process stabilizes the bit error rate at 4e-5, avoiding the compensation failure caused by sudden changes in orbit in traditional methods. In the multi-non-line-of-sight path scenario (Example 5), the algorithm ensures parameter updates under low SNR through a compression feedback mechanism, compensating for the three effective reflection paths separately, with a bit error rate of 3.5e-5, demonstrating its adaptability to complex multipath environments.

[0092] In summary, this invention significantly reduces the bit error rate in various scenarios through mechanisms such as multipath separation, formulaic frequency offset calculation, material correction, and dynamic weight adjustment, solving the problem of insufficient compensation accuracy caused by traditional methods ignoring non-line-of-sight paths.

[0093] The above embodiments can be implemented in whole or in part by software, hardware, firmware or other arbitrary combinations. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product.

[0094] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0095] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0096] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of protection of the claims.

[0097] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A dynamic optimization method for pre-compensation of Doppler frequency offset in low-Earth orbit satellites, characterized in that, Includes the following steps: (1) Multipath signal separation and path identification: Through power delay spectrum analysis of satellite received signals, identify the line-of-sight path and at least one non-line-of-sight reflection path, wherein the signal power of the non-line-of-sight reflection path is not less than -10dB of the power of the line-of-sight path; (2) Extraction of geometric parameters of reflection path: Based on the time delay information of each path in the power delay spectrum, the propagation distances corresponding to the line-of-sight path and the non-line-of-sight reflection path are inferred. Combined with the satellite orbit parameters and the real-time position information of the terminal, the direct distance between the satellite and the terminal and the total distance of the reflection path from the satellite to the reflection point and then to the terminal are calculated. (3) Doppler frequency offset decoupling calculation: Doppler frequency offset models for line-of-sight path and non-line-of-sight reflection path are constructed respectively. The frequency offset of the line-of-sight path is determined based on the radial relative velocity between the satellite and the terminal, and the frequency offset of the non-line-of-sight reflection path is determined based on the satellite velocity vector, the velocity vector of the reflection point and the unit direction vector of the reflection path. (4) Dynamic compensation parameter generation: The frequency offsets of the line-of-sight path and the non-line-of-sight reflection path are respectively used as the pre-compensation frequency offset values ​​of the corresponding paths, and the compensation weight coefficients of the two are dynamically adjusted according to the demodulation error rate fed back by the terminal, and the sum of the two coefficients is 1; the demodulation error rate fed back by the terminal is compressed and transmitted in the following way: the demodulation error rate is quantized into a 4-bit binary code, where 0000 indicates that the demodulation error rate is no greater than 1e-4, and 1111 indicates that the demodulation error rate is greater than 1e-2, and is fed back to the satellite through the control channel of the satellite downlink. The control channel adopts frequency hopping spread spectrum method for transmission to enhance anti-interference. (5) Multipath joint equalization compensation: In the terminal receiver, the frequency offset compensation amount is applied to the signals of the two paths mentioned above by the frequency domain equalizer, and the effect of multipath delay spread is eliminated by the time domain equalizer.

2. The dynamic optimization method for pre-compensation of Doppler frequency offset of low-orbit satellites as described in claim 1, characterized in that, The method for obtaining the power delay spectrum in step (1) includes: performing a fast inverse Fourier transform on the received signal to obtain the channel impulse response, and separating the delay components of the line-of-sight path and the non-line-of-sight reflection path by threshold detection, wherein the threshold is set to -15dB of the peak power.

3. The dynamic optimization method for pre-compensation of Doppler frequency offset of low-orbit satellites as described in claim 1, characterized in that, The calculation method of the velocity vector of the reflection point in step (3) includes: if the reflection point is a fixed building, the velocity vector of the reflection point is zero; if the reflection point is a moving body, the velocity vector of the reflection point is updated in real time through the motion trajectory prediction model of the reflection point reported by the terminal. The motion trajectory prediction model is constructed based on the historical location information of the reflection point collected by the terminal.

4. The dynamic optimization method for pre-compensation of Doppler frequency offset of low-orbit satellites as described in claim 1, characterized in that, The dynamic adjustment mechanism of the compensation weight coefficient in step (4) includes: when the demodulation error rate fed back by the terminal is not greater than 1e-3, the line-of-sight path compensation weight coefficient is set to 0.8 and the non-line-of-sight reflection path compensation weight coefficient is set to 0.2; when the demodulation error rate is greater than 1e-3, the non-line-of-sight reflection path compensation weight coefficient is adjusted in a way that gradually increases with the increase of the demodulation error rate, with each adjustment step not exceeding 0.1, until the demodulation error rate is not greater than 1e-2.

5. The dynamic optimization method for pre-compensation of Doppler frequency offset of low-orbit satellites as described in claim 1, characterized in that, The frequency domain equalizer in step (5) adopts the minimum mean square error criterion and determines the equalization coefficient by combining the channel frequency domain response of the line-of-sight path and the corresponding frequency offset compensation amount. The channel frequency domain response is obtained by performing a fast Fourier transform on the channel impulse response. The computational complexity of the multipath joint equalization compensation is optimized in the following way: the number of non-line-of-sight reflection paths is limited to a maximum of 3, and only reflection paths with power higher than the line-of-sight path by 20dB are compensated. The path power is determined by the peak energy of the corresponding path in the power delay spectrum.

6. The dynamic optimization method for pre-compensation of Doppler frequency offset of low-orbit satellites as described in claim 1, characterized in that, It also includes step (6): when a satellite passes overhead or the terminal motion state changes abruptly, the reflection path re-identification process is triggered, the power delay spectrum is re-estimated through short-time Fourier transform, and the total distance of the reflection path and the frequency offset of the non-line-of-sight reflection path are updated. The motion state change refers to the terminal velocity change rate exceeding a preset threshold.

7. The dynamic optimization method for pre-compensation of Doppler frequency offset of low-orbit satellites as described in claim 1, characterized in that, The demodulation error rate fed back by the terminal is compressed and transmitted in the following way: the demodulation error rate is quantized into a 4-bit binary code, where 0000 indicates that the demodulation error rate is no greater than 1e-4, and 1111 indicates that the demodulation error rate is greater than 1e-2. The code is fed back to the satellite through the control channel of the satellite downlink. The control channel uses frequency hopping spread spectrum transmission to enhance anti-interference.

8. The dynamic optimization method for pre-compensation of Doppler frequency offset of low-orbit satellites as described in claim 1, characterized in that, The method for calculating the unit direction vector of the reflection path includes: based on the three-dimensional coordinates of the satellite, the three-dimensional coordinates of the reflection point, and the three-dimensional coordinates of the terminal, the sum of the vector from the satellite to the reflection point and the vector from the reflection point to the terminal is calculated, and then the vector is normalized to obtain the unit direction vector of the reflection path.

9. The dynamic optimization method for pre-compensation of Doppler frequency offset of low-orbit satellites as described in claim 1, characterized in that, In the Doppler frequency offset decoupling calculation in step (3), for non-line-of-sight reflection paths, a frequency offset correction model based on the material attenuation coefficient of the reflection point is introduced. The material attenuation coefficient is calculated based on the dielectric constant of the reflection point and the incident angle of the electromagnetic wave. Specifically, the typical reflector material database stored in the terminal contains the dielectric constants of buildings, water bodies, and metal materials to match the reflection point type. The attenuation coefficient is dynamically adjusted in combination with the incident angle of the satellite signal. The corrected non-line-of-sight path frequency offset value = original frequency offset value × (1 - material attenuation coefficient × sinθ), where θ is the incident angle of the signal.

10. The dynamic optimization method for pre-compensation of Doppler frequency offset of low-orbit satellites as described in claim 1, characterized in that, The Doppler frequency offset decoupling calculation in step (3) uses the following formula: Line-of-sight path Doppler frequency offset calculation formula: f_d_LOS=-(v_s·u_LOS-v_t·u_LOS) / λ Where f_d_LOS is the line-of-sight path Doppler frequency offset, v_s is the satellite velocity vector, v_t is the terminal velocity vector, u_LOS is the unit direction vector from the satellite to the terminal, and λ is the signal wavelength; Formula for calculating Doppler frequency offset of non-line-of-sight reflection path: f_d_NLOS=-[v_s·u_1-v_r·(u_1-u_2)-v_t·u_2] / λ Where f_d_NLOS is the Doppler frequency offset value of the non-line-of-sight reflection path, u_1 is the unit direction vector from the satellite to the reflection point, u_2 is the unit direction vector from the reflection point to the terminal, and v_r is the velocity vector of the reflection point; The satellite velocity vector v_s is obtained through the onboard GNSS receiver, the terminal velocity vector v_t is obtained through the terminal positioning module, and the reflection point velocity vector v_r is preset according to the reflection point type or derived from the terminal motion sensor data; the unit direction vectors u_LOS, u_1, and u_2 are all obtained through three-dimensional coordinate difference vector normalization processing, and the signal wavelength λ is calculated and determined according to the satellite signal carrier frequency; in actual calculations, the vector dot product operation result of the above formula retains 6 significant decimal places, and when the reflection point is a stationary object, v_r is taken as a zero vector.

Images