Multi-user Doppler frequency offset parallel compensation method and device for on-satellite processing
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-10
- Publication Date
- 2026-03-31
Smart Images
Figure CN121770947A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of on-board processing (OBP) satellite technology, specifically a parallel Doppler frequency offset compensation method and apparatus for the demodulation module of an on-board processing type satellite receiver in a multi-user channel scenario. Background Technology
[0002] In existing technologies, as disclosed on page 1 of Xia Kewen's *Satellite Communications* (1st edition, December 2008), satellite communication refers to the technology of using artificial Earth satellites as relay stations to relay radio waves and conduct communication between two or more earth stations (also known as ground users or users). A system constructed using this technology is called a satellite communication system, where the satellite used to achieve communication is called a communication satellite, and its function is equivalent to a radio relay station deployed at extremely high altitudes. For example, as disclosed on page 43 of Zhao Long's *Satellite Communications* (1st edition, April 2022), a satellite communication system mainly consists of a space segment, a ground segment, and a control segment. The space segment mainly refers to the communication satellite itself; the ground segment mainly includes interface stations, user stations, and service stations that interact with various terminal users and the core network, collectively referred to as earth stations; the control segment is mainly responsible for status monitoring and command control of the satellite platform and communication payload, and typically consists of telemetry, remote control, and tracking stations (TTC stations) and network management stations.
[0003] In the space segment, the core function of communication satellites is signal relay. Based on their relay payload design architecture, there are two main technical paths: transparent payload and on-board processing (OBP). Satellites with a transparent payload architecture only perform frequency conversion and power amplification on the received uplink signal, without baseband demodulation or modulation. This architecture is simple and highly reliable, but it amplifies and relays noise and distortion from the uplink to the downlink, leading to noise accumulation in the end-to-end communication link. In contrast, the on-board processing architecture is more advanced. It demodulates the uplink signal on the satellite, processes the signal in the digital baseband domain, and then remodulates it before transmitting it to the ground. As disclosed on page 255 of Louis J. Ippolito's *Satellite Communications Systems Engineering* (1st edition, December 2008), a typical on-board processing transponder mainly consists of functional modules such as a receiver, demodulator, baseband processor, modulator, and power amplifier. Through the demodulation and regeneration process, this architecture can eliminate the impact of uplink noise on downlink, thereby significantly improving the end-to-end signal-to-noise ratio and achieving better communication performance.
[0004] With the rapid development of Low Earth Orbit (LEO) satellite constellations, their communication applications are becoming increasingly widespread. However, due to the high-speed motion of LEO satellites relative to earth stations (or users), significant Doppler frequency offsets occur. In multi-user access scenarios, users in different geographical locations will have different Doppler frequency offsets in their uplink signals when they reach the satellite receiver because of their different relative velocities to the satellite. This frequency offset difference between users poses a severe technical challenge to onboard processing satellite receivers, especially in their baseband preprocessing stage.
[0005] To address the Doppler frequency offset issue in onboard processing satellite receivers, existing technologies mainly employ the following two approaches:
[0006] Option 1: Joint Compensation; This option performs a unified frequency correction across the entire receiving frequency band in the RF front-end or digital domain. This correction value is typically based on a Doppler estimate of the beam center or a reference user. However, this unified compensation strategy ignores the frequency offset differences between users. When precise compensation is performed using the frequency offset of a particular user as a reference, it inevitably introduces significant residual frequency offsets to other users. These residual frequency offsets disrupt the orthogonality between user signals, thus translating into multi-user interference within the system and severely degrading communication quality. Therefore, this option cannot meet the basic frequency accuracy requirements of low-Earth orbit multi-user communication systems.
[0007] Option 2: Independent Compensation; This option performs individual frequency offset estimation and compensation for each user's signal in the digital baseband section. For example, some existing technologies use correlation operations or Fourier transforms based on the demodulation reference signal (DM-RS) to estimate the frequency offset. Theoretically, such methods can achieve accurate compensation for each user, but in practical applications, they face two major technical bottlenecks:
[0008] A) High computational complexity: The satellite needs to execute a complete and complex estimation algorithm for each user. In scenarios with a large number of users, the total computational load will increase linearly or even higher, which puts enormous pressure on the extremely limited size, weight, power consumption, and heat dissipation (SWaP) capabilities of the onboard platform.
[0009] B) The contradiction between dynamic adaptability and resource overhead: When low-Earth orbit satellites pass overhead, the Doppler frequency offset changes at an extremely rapid rate, requiring the compensation algorithm to have extremely high tracking capabilities. However, existing high-precision tracking algorithms usually rely on user terminals to send denser or higher-power reference signals, which directly leads to a large amount of valuable spectrum resources being occupied by signaling overhead, reducing the overall data transmission efficiency of the system.
[0010] In summary, existing Doppler compensation techniques in on-board processing satellite receivers either suffer from poor performance due to their inability to handle frequency offset differences between users, or are difficult to deploy effectively on resource-constrained on-board platforms due to excessive computational complexity and resource overhead. Therefore, there is an urgent need in this field to develop a novel baseband preprocessing algorithm for on-board processing satellite receivers that combines low complexity, high accuracy, and resource conservation to overcome the shortcomings of the existing technologies. Summary of the Invention
[0011] One objective of this invention is to design a Doppler frequency offset compensation device suitable for OBP (On-Board Receiver), which consists of a multi-channel signal separation unit (100), a Doppler frequency offset estimation unit (200), and a Doppler parallel compensation unit (300). This device is installed in the demodulation module of an on-board processing type low-Earth orbit satellite receiver. The Doppler frequency offset compensation device is used, on the one hand, to receive the broadband digital signals (i.e.,...) from all user channels in the demodulation module. On the other hand, it is used to output compensated and separated user data (i.e. The improved OBP satellite framework is provided to the demodulation module. Figure 1 As shown in Figure 9.8, section 9.2 On-Board Processing (OBP) Satellite, on page 255 of Louis J. Ippolito's *Satellite Communications Systems Engineering* (1st edition, December 2008). This invention addresses the technical problem of large and varied Doppler frequency offsets in uplink multi-user channels caused by the high-speed motion of low-Earth orbit satellites, leading to severe inter-user interference and degraded multi-channel communication quality. By transforming the complex joint compensation problem into a simple, parallel compensation problem, this invention significantly reduces algorithm complexity, making it highly suitable for resource-constrained onboard platforms. It also ensures the accuracy of compensation for each channel, improves the reliability and spectral efficiency of OBP satellite communication, and possesses excellent scalability.
[0012] To address the technical problem of high computational complexity in the multi-user signal coupling and channel separation stages of onboard processing satellite receivers, this invention first designs a multi-channel signal separation unit (100). Its purpose is to separate the broadband digital signal (i.e., the signal from all user channels) received on the OBP satellite receiver. In the digital domain, it is decomposed into multiple independent, narrowband sub-channel data streams (i.e. This reduces the need for complex calculations on the entire broadband signal, laying the foundation for subsequent independent and precise compensation, thus solving the bottleneck of the rapid increase in computational load caused by the increase in the number of users in existing technologies.
[0013] To address the technical problem of multi-user Doppler frequency offset compensation in satellite receivers: After achieving user sub-channel separation, this invention designs an independent frequency offset estimation stage based on a cyclic prefix, namely, a Doppler frequency offset estimation unit (200). Its purpose is to utilize the special structure of the cyclic prefix in each user channel signal to perform high-precision frequency offset estimation (i.e.,...) within their respective independent narrowband sub-channels. The calculation of the Doppler frequency offset is equivalent to calculating the relative velocity of each satellite user on the satellite receiver. Since the signal interference within each sub-channel is low and the computational load is controllable, a more refined estimation method can be used to obtain the true and accurate Doppler frequency offset value for each user. This solves the fundamental flaw of existing technologies that cannot distinguish user differences in joint compensation.
[0014] To address the technical problem of mutual interference caused by different Doppler effects in multi-user signals, this invention designs an independent compensation stage for each sub-channel, namely a parallel Doppler compensation unit (300), after accurately obtaining the frequency offset value for each user. Its purpose is to apply a digital frequency shift equal in magnitude and opposite in direction to the frequency offset value of that user channel to the data stream of each sub-channel. Through this precise compensation, the Doppler frequency offset caused by the relative velocity between the satellite and the earth station can be compensated, correcting the signal of each user back to its zero intermediate frequency position, thereby ensuring the stability and reliability of the entire communication system.
[0015] In summary, in order to address the technical challenges of excessively high computational complexity, insufficient compensation accuracy, and difficulty in deployment on resource-constrained satellite platforms, existing on-board processing satellite receivers in multi-user scenarios for Doppler frequency shift compensation methods, this invention proposes a parallel Doppler frequency shift compensation method and device for low-Earth orbit satellite multi-user communication scenarios.
[0016] The second objective of this invention is to propose a parallel Doppler frequency offset compensation method suitable for OBP (On-Board Platform). This method aims to solve the technical problem that the high-speed motion of low-Earth orbit satellites causes large and varied Doppler frequency offsets in uplink multi-user signals, leading to severe inter-user interference and degraded communication quality. To overcome the shortcomings of existing independent compensation methods, such as high computational complexity, high resource consumption, and poor scalability, this invention proposes an efficient processing scheme. The parallel Doppler frequency offset compensation method of this invention includes: first, using an efficient filter bank to decompose the received broadband multi-user signal into multiple independent narrowband sub-channel signals; then, within each narrowband sub-channel, using the cyclic prefix in the signal, independently and accurately estimating the frequency offset value unique to each user; finally, based on the frequency offset estimate, applying a reverse digital frequency shift to each sub-channel signal to complete accurate compensation. This invention transforms the complex joint compensation problem into a simple parallel compensation problem, which not only significantly reduces the algorithm complexity, making it highly applicable to on-board resource-constrained platforms, but also ensures the accuracy of compensation for each user, improves the reliability and spectral efficiency of system communication, and possesses excellent scalability.
[0017] Compared with the prior art, the method proposed in this invention has the following significant technical effects:
[0018] (1) The computational complexity is greatly reduced and it is highly suitable for on-board platforms: The present invention adopts an independent compensation strategy to decompose large-scale joint processing problems into a large number of parallel simple processing problems. Its overall computational complexity is far lower than that of traditional methods that run complex algorithms independently for each user, and it can well adapt to the strict limitations of size, weight and power consumption of low-Earth orbit satellite payloads.
[0019] (2) High compensation accuracy and good user isolation: Since each user has been independently and accurately estimated and compensated for frequency offset, this method fundamentally eliminates the differential Doppler effect between users, avoids the residual frequency offset and interference between users caused by "average compensation", and greatly improves the signal demodulation performance and system reliability.
[0020] (3) The system has strong scalability and flexible design: The device architecture of the present invention has excellent scalability. When the number of users that the system needs to support increases, only the number of parallel sub-channel processing paths needs to be increased, and the core device does not need to be rebuilt. At the same time, the method is not sensitive to the modulation method and code rate adopted by the user, and has good versatility and flexibility.
[0021] (4) Spectral efficiency is effectively improved: Due to the accurate compensation of the Doppler effect, the interference level in the system is significantly reduced and the signal-to-noise ratio is improved. This allows the system to use more efficient high-order modulation and coding schemes (such as 64-QAM, 256-QAM), or to obtain a lower bit error rate under the same conditions, thereby transmitting more data with limited satellite bandwidth resources, effectively improving the spectral efficiency and capacity of the entire satellite communication system. Attached Figure Description
[0022] Figure 1 This is a structural diagram of the fusion of the parallel Doppler frequency offset compensation device of the present invention with the demodulation module in an on-board processing type satellite receiver.
[0023] Figure 2 This is the overall power spectral density distribution diagram after the superposition of 8 user signals in the simulation of the embodiment.
[0024] Figure 3 This is an analysis graph of the Doppler frequency offset and estimated values under a 10dB signal-to-noise ratio simulation in the example.
[0025] Figure 4 This is an analysis graph of the Doppler frequency offset and estimated values under a simulation of the embodiment at a signal-to-noise ratio of 15dB.
[0026] Figure 5 This is an analysis graph of the Doppler frequency offset and estimated values under a simulation of the example at a signal-to-noise ratio of 20dB.
[0027] Figure 6 This is an error analysis diagram of Doppler frequency offset estimation under different signal-to-noise ratios in the simulation of the example.
[0028] Figure 7 This is an analysis diagram of the vector error of the QPSK constellation diagram under different signal-to-noise ratios in the simulation of the example. Detailed Implementation
[0029] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. The examples of the parameters listed are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
[0030] This invention applies to low-Earth orbit satellite receivers of the On-Board Processing (OBP) satellite type, wherein the receiver simultaneously receives signals from multiple independent users. For ease of explanation, any one user is labeled as... The total number of all users that the receiving end can receive is denoted as ,and In this invention, the total number of users It also represents the total number of channels for information communication and interaction, while the user It also represents the first Road channel. Located at the... The channel preceding the current channel is called the previous channel, denoted as . Located at the The channel following a path channel is called the next channel, denoted as . .
[0031] The OBP communication process used in this invention involves the design of system sampling rate and total bandwidth. Assuming the receiver's RF front-end down-converts and digitizes the collected uplink signals from all users, the sampling rate of the resulting broadband digital signal is denoted as... The sampling rate The corresponding spectrum range can cover all users (i.e.) Signal reception for (each user).
[0032] In the OBP communication process applied in this invention, if there is There are [number] users, each assigned an independent user channel. The average spectrum bandwidth allocated to each user is denoted as [specification]. ( This is simply referred to as the subchannel sampling rate. In the digital frequency domain, it is related to the sampling rate. The corresponding normalized digital angular frequency range is or Generally The value is 3.14. Correspondingly, the frequency width of the normalized digital angular frequency range allocated to each user is... .
[0033] In the OBP communication process applied in this invention, if there is One user, any one user (in The time-domain complex signal sequence to be transmitted generated in baseband is denoted as... . In This indicates a time-domain index identifier for low sampling rates.
[0034] In the OBP communication process applied in this invention, all Each user's signal is transmitted through their respective channels, and the summation is captured by the receiver. After radio frequency processing and analog-to-digital conversion, the sampling rate formed at the receiver is... The broadband digital signal sequence is denoted as ( (abbreviated as Broadband Digital Information), It serves as the input for all digital signal processing. In The time-domain index identifier representing the high-speed sampling rate.
[0035] In the OBP communication process applied in this invention, the After processing by the multi-channel signal separation unit 100 of the present invention, the separated signal will be... Sub-channel information corresponding to each user, i.e., signal flow, based on The signal flows of individual users form a signal flow set. Among them, when dealing with Sub-channel packet processing is performed to obtain the first The preprocessed data of each sub-channel is denoted as After filtering and separation, the resulting... The baseband output signal sequence of each sub-channel is denoted as (Right now (Simply referred to as narrowband sub-channel information). It is worth noting that... Time-domain index Corresponding to high sampling rate ,and Time-domain index Corresponding to a lower sub-channel sampling rate .
[0036] The high-efficiency filter bank used in this invention is designed based on a prototype low-pass filter, the time-domain impulse response coefficient sequence of which is denoted as... The performance of this prototype filter is determined by its bandwidth parameters, where the passband cutoff digital angular frequency is denoted as . The stopband starting digital angular frequency is denoted as To meet the above design requirements, the filter's order needs to be constrained. Let the filter's order be denoted as... .
[0037] By modifying the prototype filter coefficients The polyphase decomposition of the path channel can yield... A multiphase sub-filter. Among them, the first... Multiphase sub-filters of the channel (where) The coefficient sequence of ) is denoted as .
[0038] In the OBP communication system used in this invention, users employ Orthogonal Frequency Division Multiplexing (OFDM) signals for transmission. The generation of OFDM symbols is based on... The number of cyclic prefix points used for each OFDM symbol is denoted as: Inverse Fast Fourier Transform (IFFT) and Fast Fourier Transform (FFT). .
[0039] In the OBP communication system applied in this invention, independent Doppler frequency offset correction is required for each subchannel. The subcarrier analog frequency spacing within a single OFDM symbol is denoted as... Its value equals the signal bandwidth per user divided by the number of OFDM modulation points. ,Right now .
[0040] No. In Orthogonal Frequency Division Multiplexing (OFDM), the total analog frequency offset caused by factors such as the Doppler effect and oscillator bias plays a decisive role in signal demodulation performance due to the normalized frequency offset relative to the subcarrier spacing. Therefore, the first... The normalized Doppler frequency offset of the path channel is denoted as ,this It is a dimensionless parameter, the Doppler frequency offset. This is the objective of the frequency estimation algorithm in the method of this invention.
[0041] To further illustrate the technical concept of this invention, the specific implementation steps of the parallel compensation method for Doppler frequency offset in baseband preprocessing of multi-user channels at the OBP satellite receiver in low-Earth orbit satellite communication, proposed in this invention, will be described in detail below with reference to preferred embodiments. It should be understood that these descriptions are for illustrative purposes only and are not intended to limit the invention in any way.
[0042] The parallel Doppler frequency offset compensation method of this invention mainly includes the following three core steps: efficient filtering, independent frequency offset estimation, and accurate frequency offset compensation. Figure 1 The structural framework of a parallel compensation device for multi-user Doppler frequency offset used for on-board processing is presented.
[0043] Step 1: High-efficiency filtering for broadband multi-user channel signal separation;
[0044] In this invention, the multi-channel signal separation unit (100) obtains the broadband digital information of all users from the demodulation model of the OBP satellite, denoted as... ; will the Efficiently decomposed into Each user channel has its own independent narrowband sub-channel signal. For example, any user The corresponding number The narrowband sub-channel signal of the path channel is denoted as .
[0045] In this invention, this step is preferably implemented using an FFT polyphase filter bank, which can be further subdivided into the following sub-steps:
[0046] Step 101: Design and decomposition of the low-pass digital filter;
[0047] In this invention, a low-pass digital filter capable of performing either pre-decimation or post-interpolation is designed. This is for wideband digital information obtained from the demodulation module. Baseband decimation is performed, which involves decimating the entire band of the bandpass signal. Refer to *Software Radio Principles and Applications*, authors: Lou Caiyi, Xu Jianliang, and Yang Xiaoniu; publisher: Beijing: Electronic Industry Press, publication date: August 2014, pages 52-53, Fundamentals of Digital Filter Design.
[0048] In this invention, the time-domain impulse response coefficient sequence of the low-pass filter is denoted as... Its key frequency domain parameters include the passband cutoff digital angular frequency. and stopband start digital angular frequency To ensure that the signals of the separated sub-channels do not interfere with each other and do not experience spectral aliasing, these parameters must meet the following conditions. ,in This represents the total number of users (or the total number of channels). Based on the specific requirements for channel isolation, design a filter order that meets this metric. The type of window function (e.g., Kaiser window, Hamming window, etc.). Preferably, to facilitate the subsequent implementation of polyphase decomposition, the filter order is... Designed for total number of users Integer multiples of.
[0049] By adjusting the low-pass filter coefficients The polyphase decomposition of the path channel can yield... There are several multiphase sub-filters. The coefficient sequence of the first channel multiphase sub-filter is denoted as... The coefficient sequence of the second channel multiphase sub-filter is denoted as... . No. The coefficient sequence of the multiphase sub-filter of the path channel is denoted as The coefficient sequence of the last channel multiphase sub-filter is denoted as... Through this decomposition, a high-order filter is equivalent to... This yields a set of coefficient sequences for the multiphase sub-filters, consisting of several parallel, lower-order sub-filters. ,and The This laid the foundation for efficient processing of Doppler frequency offset.
[0050] Step 102: Signal separation based on FFT polyphase filter bank;
[0051] Utilize the design Broadband digital information Performing a separation operation includes the following steps:
[0052] Step A. Transfer broadband digital information Perform serial-to-parallel conversion and decompose into A set of parallel signal streams. The They are fed into their respective multiphase sub-filters (the channel identifier corresponds one-to-one with the filter channel identifier). Specifically, the first... The input processed by the multiphase sub-filter of the path channel is (This is an example of one implementation method.)
[0053] Step B. The outputs of each sub-filter at the same time are processed by Fast Fourier Transform.
[0054] Step C. The output of this FFT module is the final separated result. The sequence of sub-channel signals. Path output This represents the first Narrowband signals from individual users that have been separated and downsampled.
[0055] Through the processing flow of steps A-C, this invention replaces the traditional method with a structured and efficient computation. Independent digital downconversion and filtering operations greatly reduce computational complexity.
[0056] Step 2: Independent frequency offset estimation for each sub-channel based on the cyclic prefix;
[0057] In the Doppler frequency offset estimation unit (200), the cyclic prefix contained in the signal is used to perform OFDM technology for each separated sub-channel signal. The normalized Doppler frequency offset contained therein can be accurately estimated based on Formula 1. .
[0058] .
[0059] The time-domain index of the first term summed.
[0060] This represents the number of points in the OFDM modulation.
[0061] This represents the number of cyclic prefix points in OFDM modulation.
[0062] The time-domain index for the summation of the second term.
[0063] For any user The corresponding number Narrowband sub-channel signals of the road channel, middle This indicates a time-domain index identifier for low sampling rates. For In That is Temporal index in the loop section.
[0064] for . conjugate.
[0065] This represents finding the digital angular frequency of the corresponding complex number.
[0066] In this embodiment, the Doppler frequency offset estimation unit (200) modulates the user signal using OFDM. The number of points in the OFDM modulation process is denoted as... The number of cyclic prefix points used in each OFDM modulation is denoted as . .
[0067] This step utilizes the periodicity of the received cyclic prefix. By analyzing the sub-channel signal... By performing an autocorrelation operation between the cyclic prefix portion and the end of the symbol, the phase rotation information caused by the Doppler frequency offset can be extracted. Specifically, the signal can be delayed by [a certain factor]. The phase of the result obtained by conjugate multiplication of each sample version directly corresponds to the Doppler frequency offset. Through the calculation of formula (1), it is possible to obtain the phase of each sub-channel. Independently and accurately calculate the estimated value of its normalized Doppler frequency shift. It only requires... The number of multiplication operations is small, the resource consumption is low, and the cyclic prefix is the basis for OFDM communication, which makes efficient use of the structural characteristics of OFDM symbols.
[0068] Step 3: Perform independent Doppler frequency offset parallel compensation on the sub-channels based on the frequency offset estimate;
[0069] Utilizing in the Doppler parallel compensation unit (300) The frequency of each sub-channel signal is precisely corrected according to formula (2) to obtain a frequency-offset-free signal. .
[0070] .
[0071] It represents the imaginary part of a complex number.
[0072] This indicates a time-domain index identifier for low sampling rates.
[0073] As can be seen from the signal model, the first Received signal of the sub-channel It can be modeled as a primitive unbiased signal The product of the product with a complex sinusoidal signal, the phase rotation of which represents the Doppler effect, i.e. (Due to various noises, the two are not exactly equal.) To eliminate the influence of Doppler frequency shift, it is only necessary to adjust the received signal... It can be multiplied by a complex sine correction signal with the same frequency but opposite direction. It represents the imaginary part of a complex number.
[0074] This step is for each subchannel Generate a compensation sequence Then, the compensation sequence is combined with the subchannel signal. By performing point-by-point multiplication (i.e., formula (2)), the final recovered frequency-offset-free signal is obtained. .
[0075] After this step, the Doppler interference in each sub-channel signal is effectively eliminated, and The signal data of each user has been successfully separated. Figure 4 As shown, this is the compensated and separated user data. The data can then be directly sent to their respective subsequent processes such as channel estimation, phase compensation, and constellation demapping, thereby completing the entire communication reception process.
[0076] See Figure 1 As shown, this invention designs a multi-user Doppler frequency offset parallel compensation device for on-board processing. This device is installed in the demodulation module of a low-Earth orbit satellite receiver for on-board processing, belonging to the field of on-board processing (OBP) satellite technology. The device consists of a multi-channel signal separation unit (100), a Doppler frequency offset estimation unit (200), and a Doppler parallel compensation unit (300).
[0077] Multi-channel signal separation unit (100)
[0078] The multi-channel signal separation unit (100) is used, on the one hand, to receive the broadband digital signals from all user channels in the demodulation module (i.e., On the other hand, the aforementioned Efficiently decomposed into Each is an independent narrowband sub-channel signal corresponding to a user channel.
[0079] Doppler frequency offset estimation unit (200)
[0080] The Doppler frequency offset estimation unit (200) estimates the narrowband sub-channel signal corresponding to each user channel based on... Frequency offset estimation is performed to obtain the normalized Doppler frequency offset for each user channel. .
[0081] Doppler parallel compensation unit (300)
[0082] The parallel Doppler compensation unit (300) calculates the normalized Doppler frequency offset output by the Doppler frequency offset estimation unit (200). in accordance with Frequency correction is performed to obtain the frequency-offset-free signal for each user channel. The output is sent to the demodulation module.
[0083] Example:
[0084] To address the Doppler frequency offset and multi-user interference issues in uplink communication of On-Board Processing (OBP) satellites (or low-Earth orbit satellites), a simulation system for uplink communication was built using Matlab software.
[0085] This invention utilizes computer programs running on a computer, such as those based on the Matlab (version 2023b) platform, for simulation. Matlab is a high-level technical computing language and interactive environment used for algorithm development, data visualization, data analysis, and numerical computation.
[0086] This simulation depicted simultaneous communication by eight users. The total baseband bandwidth used by the eight users was 200MHz, with each user allocated 25MHz of bandwidth, including a 5MHz transition band and an actual usable bandwidth of 20MHz. The center frequency of the overall bandwidth was 4GHz, and the center frequencies for each user were denoted as follows: (Unit is MHz), then They are: 3912.5, 3937.5, 3962.5, 3987.5, 4012.5, 4037.5, 4062.5 and 4087.5 respectively.
[0087] To simulate the Doppler effect in high-dynamic communication of low-Earth orbit satellites, different relative velocities (denoted as λ) were set between each user and the satellite. (The unit is m / s), and the Mach number used in the simulation is... If Mach numbers are randomly generated, then the relative velocity... The generated values are: 8165, -5609, 5308, 7968, -6594, 5952, -8095 and -6844.
[0088] The OFDM (Orthogonal Frequency Division Multiplexing) signal used has a total of 64-point FFT (Fourier Transform), with 52 subcarriers, including 4 pilot subcarriers and 48 data subcarriers. The subcarrier spacing is 0.3906MHz, and the modulation scheme used is QPSK. The OFDM signal cyclic prefix used has 16 points.
[0089] The digital filter used in this simulation has a passband cutoff frequency of 10MHz and a stopband start frequency of 12.5MHz, a stopband attenuation of 80dB, and a filter order of 240.
[0090] Simulation results are as follows Figures 2 to 7 As shown. Figure 2 The overall power spectral density distribution of the superimposed eight user signals in the simulation is shown. The distribution of the eight user signals in the frequency domain according to preset parameters is clearly observed in the figure, constituting the multi-user signal environment of this simulation. The Doppler frequency offset value of each user's signal is simulated based on the simulated velocity value and the signal center frequency.
[0091] Figure 3 , Figure 4 and Figure 5 The comparison between the estimated frequency offset values and the true values for each user obtained using the algorithm of this invention is shown under different signal-to-noise ratios (SNR 10dB, 15dB, 20dB). The dashed line represents the ideal estimation curve. All simulation data points are closely distributed around the ideal curve, and as the signal-to-noise ratio increases to 20dB, the estimated value basically coincides with the true value. Moreover, the frequency offset estimation error of all users is controlled within 4kHz, verifying that the algorithm of this invention has high estimation accuracy under high signal-to-noise ratio.
[0092] To verify the performance of the algorithm of this invention, this embodiment is simulated under different signal-to-noise ratio conditions, and the average relative error percentage is used as a performance index to measure the estimation accuracy. Figure 6 The method of the present invention (hereinafter referred to as cp2) is shown in comparison with the prior art. method, Methods and Performance comparison results of the methods.
[0093] Method: Schmidl TM, Cox D C. Robust frequency and timingsynchronization for OFDM[J]. IEEE transactions on communications, 2002, 45(12): 1613-1621.
[0094] Method: Equation (5.28) on page 145 of "MIMO-OFDM Wireless Communication Technology and MATLAB Implementation" (1st edition, June 2013) by Yong Soo Cho and Jaekwon Kim.
[0095] Method: Moose P H. A technique for orthogonal frequency divisionmultiplexing frequency offset correction[J]. IEEE Transactions oncommunications, 2002, 42(10): 2908-2914.
[0096] Depend on Figure 6 It can be seen that the average relative error of all methods decreases with the increase of signal-to-noise ratio. In the typical operating range of satellite communication (e.g., 10dB to 20dB signal-to-noise ratio), the average relative error of the method of this invention (cp2) can be stably maintained below 7%, demonstrating its effectiveness and robustness in the target application scenario.
[0097] With the best performance Compared to other methods, when the signal-to-noise ratio is higher than 15dB, the performance difference of the method of this invention (cp2) is less than 2%. However, it should be particularly noted that... The performance advantage of traditional methods relies on additional training symbols, while the method of this invention does not require this overhead and consumes significantly fewer computational resources. Therefore, this invention effectively reduces system overhead and computational complexity while ensuring high estimation accuracy, making it particularly suitable for application scenarios with limited communication and computing resources, such as on-board processing, and has significant practical value.
[0098] Figure 7 The diagram illustrates the relationship between the average vector magnitude (EVM) of the demodulation constellation diagram and the signal-to-noise ratio (SNR) after frequency offset compensation using different methods. At an SNR of 13 dB, the measured average EVM value is better than -10 dB. This performance meets the threshold required for reliable demodulation of QPSK signals, demonstrating that the algorithm of this invention can effectively compensate for Doppler frequency offset and ensure the accuracy of subsequent signal demodulation.
Claims
1. A multi-user Doppler frequency offset parallel compensation method for on-board processing, which is arranged in a demodulation module of a low earth orbit satellite receiver of the on-board processing type; characterized in that The steps include: Step one: high-efficiency filtering of wideband multi-user channel signal separation; The multi-channel signal separation unit (100) obtains the broadband digital information of all users from the demodulation model of the OBP satellite, denoted as... ; will the Decomposed into Each independent narrowband sub-channel signal corresponds to a different user channel, denoted as […]. ; Step two: independent frequency offset estimation of each sub-channel based on a cyclic prefix; In the Doppler frequency offset estimation unit (200), the cyclic prefix contained in the signal is used to estimate the normalized Doppler frequency offset contained in the signal accurately for each separated sub-channel signal In accordance with Formula 1, the normalized Doppler frequency offset contained in the signal is estimated accurately ; ； the time domain index of the first item of summation; Number of points for OFDM modulation; Number of cyclic prefix points for OFDM modulation; the time domain index of the second item of summation; for any user corresponding first a narrowband sub-channel signal of the road channel; in denotes a time domain index identification of a low speed sampling rate; For conjugated; representing the digital angular frequency of the corresponding complex number; Step three: independent Doppler frequency offset parallel compensation of the sub-channels based on the frequency offset estimation values; In the Doppler parallel compensation unit (300) utilizes The accurate frequency correction is made to each sub-channel signal according to formula (2), and the frequency offset-free signal is obtained ; ； Represents the imaginary part of a complex number; time domain index identifying a low speed sampling rate. 2.The method for multi-user Doppler frequency offset parallel compensation for on-board processing according to claim 1, characterized in that: The decomposition of the wideband digital information in step one is performed by an FFT polyphase filter bank. The decomposition of the wideband digital information in step one is performed by an FFT polyphase filter bank.
3. The multi-user Doppler frequency offset parallel compensation method for on-board processing according to claim 2, characterized in that: Step 101: design and decomposition of a low-pass digital filter; Let the time-domain impulse response coefficient sequence of the low-pass filter be denoted as ; the passband cutoff digital angular frequency and the stopband starting digital angular frequency ; the parameters need to satisfy the condition , wherein is the total number of user channels; the filter order and the window function type are designed to satisfy the filter By adjusting the low-pass filter coefficients The multiphase decomposition of the path channel yields There are 1 multiphase sub-filters; where the coefficient sequence of the first channel multiphase sub-filter is denoted as... The coefficient sequence of the second channel polyphase sub-filter is denoted as... ;No. The coefficient sequence of the multiphase sub-filter of the path channel is denoted as The coefficient sequence of the last channel polyphase sub-filter is denoted as... Through this decomposition, a higher-order filter is equivalent to... This yields a set of coefficient sequences for the polyphase sub-filters, consisting of several parallel, lower-order sub-filters. ,and ; Step 102: signal separation based on an FFT polyphase filter bank; Using the designed on broadband digital information performing the separation operation: Step A. Transfer broadband digital information Perform serial-to-parallel conversion and decompose into A set of parallel signal streams. The They are fed into their respective multiphase sub-filters; Step B. The outputs of each sub-filter at the same time are processed by Fast Fourier Transform; Step C. The output of this FFT module is the final separated sub-channel signal sequence; the th output represents the separated and down-sampled narrowband signal of the th user.
4. A device for implementing the method of any one of claims 1-3, the device being arranged in a demodulation module of a low earth orbit satellite receiver of the star processing type; characterized in that: The device is composed of a multi-channel signal separation unit (100), a Doppler frequency offset estimation unit (200), and a Doppler parallel compensation unit (300). The multi-channel signal separation unit (100) is used for receiving the wideband digital signals of all user channels in the demodulation module (i.e. ), and separating the signals into independent narrowband sub-channel signals corresponding to the user channels. The Doppler frequency offset estimation unit (200) performs frequency offset estimation on the narrowband subchannel signals corresponding to each user channel according to to obtain the normalized Doppler frequency offset corresponding to each user channel ; The Doppler parallel compensation unit (300) carries out frequency correction on the normalized Doppler frequency offset output by the Doppler frequency offset estimation unit (200) to obtain a frequency-offset-free signal corresponding to each user channel According to carries out frequency correction, and obtains a frequency-offset-free signal corresponding to each user channel to a demodulation module.