FSO / RF hybrid communication method based on Doppler compensation and interference suppression

Abstract

The invention discloses an FSO / RF hybrid communication method based on Doppler compensation and interference suppression, and belongs to the technical field of satellite communication. The method comprises the following steps: S1, constructing a satellite-ground transmission link topology; s2, establishing an FSO link loss model; s3, establishing a directional error geometric loss statistical model; s4, establishing an RF link loss model; s5, FSO link Doppler frequency shift compensation is carried out; and S6, performing RF link multi-user interference compensation. According to the method, a received signal is converted from a time-frequency domain to a Doppler domain, and frequency deviation of an FSO link is compensated through DDT; and the RF link user interference is further reduced by adopting EM clustering, hybrid precoding, SIC and other signal processing technologies. The whole scheme of the invention has good engineering feasibility, can be flexibly deployed between the HAP and the ground gateway station, supports link adaptive scheduling and signal processing in a high-speed dynamic environment, and significantly improves the communication efficiency and robustness of a satellite-ground hybrid link system.

Description

Technical Field

[0001] The present invention belongs to the technical field of satellite communication, and particularly relates to an FSO / RF hybrid communication method based on Doppler compensation and interference suppression. Background Art

[0002] In the space-air-ground integrated network (SAGIN), the free-space optics (FSO) / radio frequency (RF) hybrid system is considered a highly promising solution for establishing high-throughput and highly reliable links. The FSO link can solve the problem of scarce RF spectrum and improve data transmission efficiency. When the FSO link is blocked, the RF link with strong penetration ability can be used as a supplementary solution. In short, the combination of FSO and RF links creates favorable conditions for establishing high-capacity wireless connections.

[0003] During the implementation of the FSO communication link, the FSO link is restricted by various practical factors. For example, the propagation of light waves in free space is affected by environmental factors, including absorption, scattering, and diffraction effects. Beam drift and scintillation caused by changes in air temperature and pressure also affect the link performance. Currently, existing studies have addressed issues such as atmospheric turbulence, pointing error, and beam drift through non-convex optimization, adaptive optical compensation, channel coding, and space-time diversity to reduce their impact. However, these studies do not consider the Doppler effect caused by the high-speed movement of satellites. The Doppler effect is caused by the relative motion between the satellite and the ground receiver, resulting in a shift in the frequency of the received signal. The magnitude of this frequency shift depends on the relative speed. When the satellite approaches the receiver, the frequency increases; when it moves away, the frequency decreases. This frequency shift will affect the synchronization and demodulation of the communication link.

[0004] In addition, in practical applications, due to factors such as the occlusion effect and path loss, the quality of the satellite link cannot always be guaranteed. Therefore, it is difficult to use FSO direct communication for the ground-to-high altitude platform (HAP) segment link when affected by obstacles such as buildings. Some existing technologies use RF, terahertz (THz), or millimeter-wave links to replace FSO for ground-HAPs segment transmission. However, the existing solutions do not clearly describe the signal processing process after the signal passes through the hybrid link, and also rarely consider the user interference problem in the RF link from the perspective of signal processing. As a result, it is difficult to suppress interference in the RF link under high user density, which in turn affects the overall performance of the system. Summary of the Invention

[0005] Aiming at the dual challenges faced by the FSO / RF hybrid system in SAGIN in the prior art: on the one hand, the high-speed movement of satellites significantly amplifies the Doppler frequency shift in the FSO link, resulting in frequency offset and signal distortion; on the other hand, the severe user interference in the RF link greatly weakens the system performance. The present invention proposes an FSO / RF hybrid communication method based on Doppler compensation and interference suppression. A satellite-HAPs-ground link model is established. Among them, the HAPs-ground segment supports both the FSO / RF hybrid link and the pure RF link. In the FSO / RF hybrid link, an unmanned aerial vehicle (UAV) is introduced as a relay node to effectively avoid the occlusion effect of the FSO link and reduce the burden on the spectrum resources of the RF link. At the same time, to improve the alignment accuracy of the transmit beam, the present application proposes a new statistical model of geometric loss of pointing error, comprehensively considering the receiver position vibration, the transmitter position vibration, and the inclination deviation caused by satellite movement. For the FSO link, the received signal is mapped from the time-frequency domain to the delay-Doppler domain, and the Doppler frequency shift caused by the high-speed movement of the satellite is compensated through the discrete Doppler transform (DDT) to solve the Doppler frequency shift problem. For the RF link, at the transmitter end, traffic clustering is completed through expectation maximization (EM) clustering to reduce the interference to the receiver caused by the mixing of different traffic. After hybrid precoding processing, the receiver end adopts serial interference cancellation (SIC) technology to further reduce the inter-cluster interference (MCI) and intra-cluster interference (ICI), thereby improving the overall transmission efficiency and link stability of the system.

[0006] First, an FSO link loss model is modeled based on four loss parameters, with the impact of the high-speed movement of satellites on pointing error modeling being considered; an RF link loss model is modeled based on the SR model, taking into account the effects of multipath propagation and shadow fading. Then, the received signal is converted from the time-frequency domain to the Doppler domain, and the frequency offset of the FSO link is compensated through DDT; and signal processing technologies such as EM clustering, hybrid precoding, and SIC are used to further reduce the user interference in the RF link.

[0007] To solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0008] An FSO / RF hybrid communication method based on Doppler compensation and interference suppression, the method comprising the following steps:

[0009] Step S1: Construct a satellite-ground transmission link topology; adopt an FSO / RF hybrid link for the link between the high-altitude platform and the ground access point, and construct a topology structure based on the FSO / RF hybrid link;

[0010] Step S2: Establish an FSO link loss model; the FSO link loss model comprehensively considers atmospheric attenuation, atmospheric turbulence, and AoA fluctuation; atmospheric turbulence considers two atmospheric turbulence modes, LN and GG; derive the link loss caused by atmospheric attenuation, the channel loss of weak turbulence under the LN model, the channel loss of medium-strong turbulence under the GG model, and the channel loss caused by AoA fluctuation;

[0011] Step S3: Establish a pointing error geometric loss statistical model; the pointing error geometric loss statistical model comprehensively considers the position deviation of the high-altitude platform receiver, the position deviation of the satellite transmitter, and the change in the satellite transmitter inclination angle caused by the high-speed movement of the satellite, and derives the pointing error; comprehensively consider the impacts of atmospheric attenuation, atmospheric turbulence, pointing error, and AoA fluctuation to obtain the channel loss coefficient of the FSO link;

[0012] Step S4: Establish an RF link loss model, and introduce the SR model to calculate the RF link channel loss;

[0013] Step S5: FSO link Doppler frequency shift compensation; adopt binary phase shift keying modulation and erbium-doped fiber amplifier gain control at the satellite transmitter; the ground access point receiver adopts the DDT algorithm to compensate for the Doppler frequency shift caused by the high-speed movement of the satellite;

[0014] Step S6: RF link multi-user interference compensation; the satellite transmitter performs traffic clustering on the access services through the EM algorithm, and combines analog precoding and digital precoding for joint signal preprocessing; the SIC algorithm is adopted at the ground access point receiver to further suppress multi-user interference.

[0015] Furthermore, in Step S2, the link loss caused by atmospheric attenuation is expressed as:

[0016] h al =exp(-Lv)

[0017] where h al represents the link loss caused by atmospheric attenuation, v represents the attenuation coefficient, and L represents the optical link length;

[0018] The channel loss of weak turbulence under the LN model is expressed as:

[0019]

[0020] where, represents the channel loss of weak turbulence under the LN model, is the Rytov variance;

[0021]

[0022] where, Denotes the channel loss of moderate-to-strong turbulence under the GG model, where the parameters α and β represent the effective numbers of large-scale and small-scale vortices respectively, and Γ(·) is the Gamma function;

[0023] The channel loss caused by AoA fluctuations is expressed as:

[0024]

[0025] where h aoa denotes the channel loss caused by AoA fluctuations, θ FoV denotes the field of view angle of the high-altitude platform receiver, and σ angle,u denotes the combined variance in the horizontal and vertical directions of the high-altitude platform.

[0026] Furthermore, in step S3, the pointing error is expressed as:

[0027]

[0028] where h pe denotes the pointing error, A represents the fraction of energy collected, and η a 2 denotes the ratio of the square of the equivalent beam width to the total displacement variance.

[0029] Furthermore, in step S4, the RF link channel loss is expressed as:

[0030]

[0031] where P is the power of the RF link, N0 is the RF link noise power, and γ RF denotes the SNR of the RF link, denotes the average SNR of the RF link, P L denotes the average power of the LOS component, 2b0 denotes the average power of the scattered component, 1F1(·,·,·) is the confluent hypergeometric function of the first kind, and m is the shape parameter of the Nakagami-m distribution.

[0032] Furthermore, in step S5, the output signal of the satellite transmitter is:

[0033]

[0034] where b n is the signal symbol, and p(t - nT) is the pulse shaping function. T is the symbol duration, i.e., the time interval for each symbol, t - nT represents the signal within the nth symbol time interval, and n represents the symbol time interval index;

[0035] At the ground access point receiver, the signal after Doppler compensation is:

[0036]

[0037] Among them, represents the DDT transform of the link with power p n , ω'(t) represents the signal perturbation term, j represents the imaginary unit, G(l a ) represents the amplification gain, and Ω(τ) is the Doppler frequency shift related to the delay.

[0038] Compared with the prior art, the present invention has the following beneficial technical effects:

[0039] (1) For the FSO link, the present invention constructs for the first time a geometric statistical model of pointing error that simultaneously considers receiver vibration, transmitter position offset, and transmitter tilt change caused by the high-speed movement of the satellite, providing theoretical support for link loss modeling and performance evaluation.

[0040] (2) Aiming at the Doppler frequency shift problem caused by high-speed movement in the FSO link, a Doppler frequency shift compensation method based on DDT compensation is proposed, effectively improving the signal demodulation performance and link stability in a dynamic environment.

[0041] (3) In the RF link part, the present invention innovatively introduces a traffic clustering strategy based on the EM algorithm, combined with analog and digital hybrid precoding and the ZF algorithm, significantly suppressing ICI and MCI. At the same time, a SIC and dynamic priority decoding process is designed at the receiving end to achieve efficient scheduling of multi-user link resources and improve system throughput and single-user SNR.

[0042] (4) The overall solution has good engineering feasibility and can be flexibly deployed between HAPs and ground gateways, supporting link adaptive scheduling and signal processing in a high-speed dynamic environment, significantly improving the communication efficiency and robustness of the satellite-ground hybrid link system. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0044] Figure 1 is a schematic diagram of the FSO communication link supported by the HAPs of the present invention.

[0045] Figure 2 is a schematic diagram of the satellite-ground transmission link of the present invention.

[0046] Figure 3 is a schematic diagram of the AoA fluctuation of the present invention.

[0047] Figure 4 It is a schematic diagram of the Gaussian beam footprint at the receiver aperture of the present invention.

[0048] Figure 5 It is a schematic diagram of the principle of the optical downlink of the present invention. Specific embodiments

[0049] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative work shall fall within the protection scope of the present invention.

[0050] The present invention proposes an FSO / RF hybrid communication method based on Doppler compensation and interference suppression. The method includes the following steps:

[0051] Step S1: Construct a space-ground transmission link topology.

[0052] As Figure 1-2 shown, the link between the high-altitude platform and the ground access point (Access Point, AP) adopts an FSO / RF hybrid link, and a topology structure based on the FSO / RF hybrid link is constructed. The FSO / RF hybrid link uses an unmanned aerial vehicle (Unmanned Aerial Vehicle, UAV) as a relay node, where the FSO link is used as the main link and the RF link is used as a backup link. The deployment height of the UAV is set to be between 1 and 2 kilometers above the ground. There are no buildings or mountain blocks in this height area, so as to ensure that the FSO link established above this height is not affected by the masking effect.

[0053] Step S2: Establish an FSO link loss model. The FSO link loss model comprehensively considers factors such as atmospheric attenuation, atmospheric turbulence, and AoA fluctuation, derives the mathematical expression of the link loss, and establishes the corresponding probability statistical model, providing a theoretical basis for subsequent link performance evaluation.

[0054] (1) Atmospheric attenuation

[0055] First, consider the atmospheric attenuation in space-ground communication. Its main sources include molecular absorption, aerosol scattering, and Rayleigh scattering. These factors will significantly affect the signal quality as the transmission frequency, meteorological conditions, and the thickness of the atmosphere along the propagation path change. For an optical link with a length of L, the attenuation loss is expressed by the Beer-Lambert law as:

[0056] h al = exp(-Lξ)(1)

[0057] where hal Denotes the link loss caused by atmospheric attenuation. ξ is the attenuation coefficient and can be expressed as λ is the optical wavelength of the FSO link, V i is the visibility, and q(V i ) is the atmospheric attenuation visibility coefficient, expressed as:

[0058]

[0059] To analyze the statistical characteristics of the path loss, it is necessary to clarify the link loss h al caused by atmospheric attenuation. The probability density function (PDF) of h is required to support subsequent signal processing derivations. In Equation (1), since the attenuation coefficient ξ follows a normal distribution N(μ, σ 2 ) with an expectation of μ and a variance of σ 2 , then -Lξ follows a normal distribution N(-Lμ, L 2 σ 2 ) with an expectation of -Lμ and a variance of L 2 σ 2 ). Therefore, logh al also follows a normal distribution N(-Lμ, L 2 σ 2 ) with an expectation of -Lμ and a variance of L 2 σ 2 ). According to the properties of the lognormal distribution, the probability density function of the link loss h al caused by atmospheric attenuation is:

[0060]

[0061] where denotes the probability density function of the link loss h al caused by atmospheric attenuation.

[0062] (2) Atmospheric Turbulence

[0063] Next, consider the impact of atmospheric turbulence in satellite-ground communication. Atmospheric turbulence is caused by random fluctuations in temperature, pressure, and refractive index in the air, which can lead to wavefront distortion and intensity scintillation during the propagation of light waves, thereby significantly affecting the communication performance of satellite-ground links. To simulate the fading caused by atmospheric turbulence, two atmospheric turbulence models, namely lognormal (LN) and gamma-gamma (GG), are considered. The LN model is applicable to weak turbulence conditions, while the GG model is applicable to medium and strong atmospheric turbulence conditions.

[0064] To derive the statistical characteristics of satellite-ground link loss, first consider weak turbulence conditions. Based on the LN model, the probability density function of the channel loss under weak turbulence is:

[0065]

[0066] Among them, represents the channel loss of weak turbulence under the LN model of the probability density function, represents the channel loss of weak turbulence under the LN model, is the Rytov variance. Since the satellite transmitter and the ground receiver are at different altitudes, the Rytov variance is expressed as:

[0067]

[0068] Among them, H is the altitude of the satellite transmitter, h0 is the altitude of the high-altitude platform, l is the altitude, ζ is the satellite zenith angle, and is the refractive index structure constant, and V ω is the root-mean-square wind speed, S t is the nominal value of the ground refractive index structure constant of.

[0069] It can be obtained from Equation (4) that the channel loss of weak turbulence based on the LN model is expressed as:

[0070]

[0071] Among them, represents the integrand. Since this integral is an unresolvable integral, the Laplace approximation is used to calculate this integral. To find the extreme point of the integrand , the form after taking the logarithm of Equation (6) is:

[0072]

[0073] For take the derivative to find the point where the derivative is zero is the integrand of the extreme point. According to the Laplace approximation method, near , the integrand can be approximated by a quadratic Taylor expansion, that is:

[0074]

[0075] Among them, represents the extreme point of the channel loss of weak turbulence under the LN model of the probability density function.

[0076] Secondly, consider the medium-strong turbulence condition. The probability density function of the channel loss of medium-strong turbulence based on the GG model is:

[0077]

[0078] Among them, represents the channel loss of moderate-to-strong turbulence under the GG model, represents the channel loss of moderate-to-strong turbulence under the GG model is the probability density function, and Γ(·) is the Gamma function, represents the modified Bessel function of the second kind of order α-β. The parameters α and β represent the effective numbers of large-scale and small-scale vortices respectively. Considering the influence of beam drift, the parameters α and β of the untracked collimated beam are defined as:

[0079]

[0080] Similarly, based on the channel loss of moderate-to-strong turbulence under the GG model is expressed as:

[0081]

[0082] (3) AoA fluctuation

[0083] When the high-altitude platform is hovering, affected by wind force or its own stabilization system, its attitude may change slightly, resulting in the deviation of the laser beam angle reaching the high-altitude platform receiver and generating AoA fluctuation. The influence of AoA fluctuation on beam reception is as Figure 3 shown. Due to the influence of AoA fluctuation, the beam is no longer orthogonal to the receiving plane. The following is the process of solving the probability density distribution function of the AoA link loss. Similar to the above, based on the probability density distribution function of the AoA link loss, the overall statistical characteristics of the satellite-ground link loss can be deduced.

[0084] The AoA deviation angle θ of the FSO link between the satellite and the high-altitude platform a is:

[0085]

[0086] In the formula, θ rx and θ ry represent the horizontal and vertical direction deviations of the high-altitude platform. Since the vibration of the high-altitude platform receiver is caused by many random events, θ rx and θ ry follow a Gaussian distribution, and the variances are and Therefore, θ a follows a Rayleigh distribution, and the probability density function of θ a is:

[0087]

[0088] Among them, Denote the AoA deviation angle as θ a of the probability density function, and σ angle,u denotes and the joint variance, and the impact of AoA fluctuations on satellite-ground transmission can be expressed as the AoA deviation angle θ a and the field of view angle θ FoV of the high-altitude platform receiver.

[0089] Consider whether the incident laser is within the receiving field of view as a 0-1 distribution event. Mathematically, given a fixed field of view, if the AoA deviation angle θ a is greater than the field of view angle θ FoV of the high-altitude platform receiver, then a link interruption occurs (i.e., h aoa = 0, where h aoa represents the link loss caused by AoA fluctuations), otherwise the maximum signal power is received (i.e., h aoa = 1). Therefore, the probability density function corresponding to h aoa is:

[0090]

[0091] where, represents the square of the field of view angle θ FoV of the high-altitude platform receiver, and δ(·) is the Dirac function.

[0092] According to Equation (14), the channel loss h aoa caused by AoA fluctuations is:

[0093]

[0094] Since δ(h aoa ) is non-zero only when h aoa = 0, the integral result of the first δ function term in Equation (14) is 0; the second δ function term δ(h aoa - 1) in Equation (14) is non-zero only when h aoa = 1. If h aoa = 1, the integral of the second term is

[0095] Step S3: Establish a statistical model for pointing error geometric loss.

[0096] The pointing error geometric loss statistical model comprehensively considers the position deviation of the high-altitude platform receiver, the position deviation of the satellite transmitter, and the change in the satellite transmitter inclination angle caused by the high-speed movement of the satellite. Based on the chi-square distribution, it derives the statistical characteristics of the radial displacement, establishes the geometric loss distribution function, and quantifies the impact of the pointing error on the link performance. Finally, by combining atmospheric attenuation, atmospheric turbulence, pointing error, and AoA fluctuations, it quantifies the relationship between the total link loss and the transmission distance, atmospheric conditions, and transmission system parameters.

[0097] It is considered that the pointing error is affected by the random position change of the high-altitude platform receiver, the random position change of the satellite transmitter, and the inclination angle of the satellite transmitter. As Figure 4 shown, the total radial displacement vector from the receiving aperture center of the high-altitude platform receiver to the beam center is denoted as r d = [x d , y d , where x d represents the x-axis value of the total radial displacement, and y d represents the y-axis value of the total radial displacement. The total radial displacement vector from the receiving aperture center of the high-altitude platform receiver to the beam center consists of the following three different vectors: i) the displacement vector r t caused by the position deviation of the satellite transmitter, r t = [x t , y t , where x t represents the x-axis value of the displacement vector caused by the position deviation of the satellite transmitter, and y t represents the y-axis value of the displacement vector caused by the position deviation of the satellite transmitter; ii) the displacement vector r r caused by the position deviation of the high-altitude platform receiver, r r = [x r , y r , where x r represents the x-axis value of the displacement vector caused by the position deviation of the high-altitude platform receiver, and y r represents the y-axis value of the displacement vector caused by the position deviation of the high-altitude platform receiver; iii) the displacement vector caused by the satellite transmitter inclination deviation Based on the above definitions, the process of deriving the probability density function of the link loss caused by the pointing error is as follows.

[0098] To calculate the link loss caused by the pointing error, it is first necessary to calculate the probability density function of the elevation angle change under the relative movement of the satellite and the ground access point. According to the central limit theorem, the position deviation is caused by many random events and follows a Gaussian distribution. Therefore, the displacement vectors r t caused by the position deviation of the satellite transmitter and r r caused by the position deviation of the high-altitude platform receiver are quantified as Gaussian distributions with a mean of 0:

[0099]

[0100] Among them, represents the displacement vector r caused by the position deviation of the satellite transmitter t which is the variance of the Gaussian distribution that r follows, represents the displacement vector r caused by the position deviation of the high-altitude platform receiver r which is the variance of the Gaussian distribution that r follows.

[0101] Different from previous studies, in the present invention, the displacement vector caused by the inclination deviation of the satellite transmitter is established as the elevation change caused by the satellite movement. Figure 5 For the schematic diagram of the optical downlink considered, the link parameters are defined as follows. θ div is the divergence angle of the satellite transmitter, dp is the displacement generated by the satellite's movement within time t, V orb is the operating speed of the satellite along the orbit, β0 is the satellite elevation angle, and S'(t) is the final position where the satellite moves from its original position within time t.

[0102] In the orbital plane, the operating speed of the satellite along the orbit can be expressed as V orb =(v r , v θ ) in the orbital plane coordinate system. is the radial velocity component, is the tangential velocity component, ν is the mean anomaly of the satellite, representing the instantaneous position of the satellite on the orbit. g is the gravitational constant of the earth, e is the orbital eccentricity, ω d is the specific angular momentum of the satellite.

[0103] The operating speed V of the satellite along the orbit orb =(ν r , v θ ) can be converted to the velocity vector V in the geocentric equatorial coordinate system through the rotation matrix R GCI :

[0104] V GCI =R·V orb (17)

[0105] Among them Ω is the longitude of the ascending node, θ i is the orbital inclination, and ω is the argument of perigee.

[0106] Assuming the coordinates of the satellite are (S x (t), S y (t), S z (t)), then the position of the satellite after movement (S' x (t), S' y (t), S' z(t)) is:

[0107]

[0108] Therefore, the total spatial position offset vector d caused by satellite motion s is (S x (t) - S' x (t), S y (t) - S' y (t), S z (t) - S' z (t)). Considering the influence of the elevation angle between the satellite and the ground, the deviation vector d at the ground access point receiver p is expressed as:

[0109]

[0110] where d x represents the spatial position offset vector along the x-axis caused by satellite motion, d y represents the spatial position offset vector along the y-axis caused by satellite motion, d z represents the spatial position offset vector along the z-axis caused by satellite motion, Δβ0 is the change in elevation angle within time t, and d s is the total spatial position offset vector caused by satellite motion.

[0111] According to Equation (18), the magnitude ||d s || of the total spatial position offset vector d caused by satellite motion is: s || is:

[0112]

[0113] where i represents the imaginary unit.

[0114] From Equation (19), obtaining the change in elevation angle Δβ0 within time t also requires the value of d p . d p follows a Gaussian distribution and has the following mean and covariance matrix:

[0115]

[0116] where represents the variance of the Gaussian distribution that d p follows along the x-axis, represents the variance of the Gaussian distribution that d p follows along the y-axis, represents the variance of the Gaussian distribution that d p follows along the z-axis.

[0117] For the sum of squares of normally distributed random variables, the result follows a chi-square distribution:

[0118]

[0119] where χ 2 (3) is a chi-square distribution with 3 degrees of freedom.

[0120] The distribution of the norm ||d p || of the deviation vector can be further derived from the properties of the chi-square distribution:

[0121]

[0122] where is the square root of a chi-square distribution with 3 degrees of freedom, and its distribution is close to the generalized Laplace distribution.

[0123] Approximating the value of ||d p || with the expected value of ||d p ||, the expected value E[||d p ||] of ||d p || is:

[0124]

[0125] Combining Equation (20) and Equation (24) to obtain cosΔβ as:

[0126]

[0127] The total radial displacement vector from the center of the receiving aperture of the high-altitude platform receiver to the beam center can be expressed as where Therefore, r d is the square root of the sum of squares of four independent normally distributed random variables with a mean of 0. It follows a chi-square distribution with 4 degrees of freedom. The probability density function of the total radial displacement vector r d from the center of the receiving aperture of the high-altitude platform receiver to the beam center is:

[0128]

[0129] where, represents the total displacement variance, and it can be solved from the above equation as:

[0130]

[0131] where erf(·) represents the error function, and the total displacement variance is:

[0132]

[0133] Therefore, the pointing error h pe has a probability density function that can be written as:

[0134]

[0135] where A = [erf(γ)] 2 is the fraction of harvested energy, γ represents the link normalized power harvesting coefficient, ξ is the atmospheric attenuation coefficient, r a is the aperture radius of the high-altitude platform receiver, ω z is the beam width of the beam at the high-altitude platform receiver. is the ratio of the square of the equivalent beam width to the total displacement variance, and the equivalent beam width

[0136] From Equation (29), considering the contribution of the error in a small range and setting the integration interval to [0, 1], the pointing error can be obtained as:

[0137]

[0138] Finally, considering the combined effects of atmospheric attenuation, atmospheric turbulence, pointing error, and AoA fluctuations, the channel loss coefficient h FSO of the FSO link and its probability density function are respectively:

[0139] The channel loss coefficient h FSO of the FSO link under the GG model and its probability density function are respectively:

[0140]

[0141] The channel loss coefficient h FSO of the FSO link under the LN model and its probability density function are respectively:

[0142]

[0143]

[0144] Step S4: Establish an RF link loss model.

[0145] When the high-altitude platform flies at different altitudes, it may experience a transition from line-of-sight (LOS) to non-line-of-sight (NLOS). The present invention proposes a link loss model based on the SR model, introducing the SR distribution model to calculate the performance of the RF link. This model combines the Nakagami-m distribution and the log-normal distribution to dynamically reflect the effects of multipath and shadow fading.

[0146] In the SR model, the probability distribution function is modeled by the Nakagami-m distribution. m is the shape parameter of the Nakagami-m distribution, which controls the severity of fading: m = 0 applies to urban areas where the direct path is blocked, and 0 < m < ∞ applies to rural and suburban environmental conditions where the LOS is partially blocked. In addition, when m = ∞, the LOS signal is unobstructed in open areas. By setting m = 0 and m = ∞, the SR model can be simplified to the lognormal distribution and the Rayleigh distribution.

[0147] The probability density function of the signal-to-noise ratio SNR of the SR fading channel in the RF link can be expressed as:

[0148]

[0149] where γ RF represents the SNR of the RF link, represents the average SNR of the RF link, P L represents the average power of the LOS component, 2b0 represents the average power of the scattered component, and s represents the line-of-sight / non-line-of-sight state indicator variable. Among them, 1F1(·,·,·) is the confluent hypergeometric function of the first kind.

[0150] The SNR γ of the RF link RF and the RF link channel loss h RF are related as follows:

[0151]

[0152] where P is the power of the RF link and N0 is the RF link noise power.

[0153] According to Equation (34) and Equation (32), the probability density function of h RF can be obtained as:

[0154]

[0155] where, represents the probability density function of the SNR of the SR fading channel in the RF link.

[0156] Finally, the RF link channel loss h RF can be expressed as:

[0157]

[0158] Since this integral contains hypergeometric functions, it can only be solved numerically after substituting specific values.

[0159] Step S5: Doppler frequency shift compensation for the FSO link. Binary Phase Shift Keying (BPSK) modulation and Erbium Doped Fiber Amplifier (EDFA) gain control are adopted in the satellite transmitter; the DDT algorithm is adopted in the ground access point receiver to compensate for the Doppler frequency shift caused by the high-speed satellite movement, ensure the signal frequency stability and synchronization, achieve stable received signal frequency, and enhance synchronization.

[0160] The FSO link transmitted signal is modulated using BPSK, and the output signal of the satellite transmitter is:

[0161]

[0162] where b n is the signal symbol, and p(t - nT) is the pulse shaping function. T is the symbol duration, that is, the time interval of each symbol. t - nT represents the signal within the nth symbol time interval, and n represents the symbol time gap index.

[0163] The modulated optical signal is amplified by the EDFA. The amplification gain G(l a ) is:

[0164]

[0165] where σ a is the absorption cross section, representing the ability of erbium ions to absorb pump light, and σ e is the emission cross section, representing the ability of erbium ions to emit stimulated radiation photons. N sr is the doping concentration of erbium ions in the optical fiber, l a is the length of the EDFA active optical fiber, and n sp is the spontaneous emission factor, which represents the intensity of spontaneous emission noise.

[0166] The signal undergoes optoelectronic conversion at the ground access point receiver to obtain the instantaneous SNR of the ground received signal:

[0167]

[0168] where γ FSO represents the SNR of the FSO link, η is the optoelectronic conversion efficiency, τ = 1 is the heterodyne detection coefficient, represents the noise power of the ground access point receiver.

[0169] The ground received signal is mapped from the time-frequency domain to the delay-Doppler domain, and the channel response can be expressed as:

[0170] h(τ, t) = g(τ)e -jΩ(τ)t (40)

[0171] Among them, \(g(\tau)\) is the amplitude response of the signal at each delay, \(\Omega(\tau)\) is the Doppler frequency shift related to the delay, and \(j\) represents the imaginary unit.

[0172] The output signal \(s(t)\) of the satellite transmitter is mapped from the time-frequency domain to the delay-Doppler domain, and the ground received signal at time \(t\) is:

[0173]

[0174] Among them, \(s(t - \tau)\) represents the waveform of the output signal of the satellite transmitter at a delay of \(\tau\), is additive white Gaussian noise, The DDT of is defined as:

[0175]

[0176] Among them, \(g(t - \tau)\) represents a window function centered at \(\tau\), that is, the variables within the range of \(t\) centered at \(\tau\) of \(g(t)\) are studied. When dealing with signals or data of finite length, directly performing Fourier transform will introduce unwanted spectral components. The window function weights the signal to gradually reduce it at the boundaries of the signal, thereby reducing spectral distortion caused by truncation.

[0177] At the ground access point receiver, the signal after Doppler compensation is:

[0178]

[0179] Among them represents performing DDT transform on the link with power \(p\) n , \(\omega'(t)\) represents the signal perturbation term,

[0180] Step S6: RF link multi-user interference compensation. The satellite transmitter performs traffic clustering on the access services through the EM algorithm, divides the services into three categories: enhanced mobile broadband service (eMBB), ultra-reliable low-latency communication service (URLLC), and massive machine-type communication service (mMTC), and combines analog precoding and digital precoding for joint signal preprocessing to reduce intra-cluster and inter-cluster interference.

[0181] At the ground access point receiver, the SIC algorithm is used to further suppress multi-user interference and improve the link transmission performance.

[0182] In the RF link, to reduce interference between different service types, traffic clustering is first performed on the access traffic. The feature vector of each service flow is defined as:

[0183] X i =[B i ,T i ,Ri , D i (44)

[0184] where B i is the bandwidth requirement, T i is the latency requirement, R i is the reliability requirement, D i is the connection density.

[0185] First, the K-means clustering is adopted to obtain the initial classification result, and then the EM algorithm is used to maximize the log-likelihood function of the Gaussian mixture model:

[0186]

[0187] where P(Y, Z|Θ) represents the joint probability distribution of the observed data Y and the latent variable Z under the parameter Θ, Y represents the observed data set, that is, all input samples {x1, x2,..., x n}, Z represents the latent variable (the class indicator variable indicating which mixture component each sample belongs to), and Θ represents the set of GMM model parameters, including N represents the total number of observed samples, π k is the mixing coefficient, N(X i |μ k , Σ k ) represents the probability density function value of the k-th Gaussian distribution at the point X i , μ k is the mean matrix, Σ k is the covariance matrix, k represents the class serial number, and K represents the total number of classes.

[0188] After the clustering is completed, according to the service requirements of each class, analog precoding and digital precoding are applied respectively. The analog precoding is implemented by using a Q-level quantization phase shifter, and the phase calculation formula is:

[0189]

[0190] where ω' represents the phase of the satellite transmitter after analog precoding, Q represents the phase quantization accuracy, ω p represents the p-th candidate phase quantization level, and angle{h RF} represents the phase angle of the RF link channel loss h RF .

[0191] The analog precoding matrix is expressed as:

[0192]

[0193] where N TLet \(N\) represent the number of satellite transmitter antennas, and \(k'\) represent the phase index (sub - carrier index) for quantization.

[0194] Digital precoding adopts the ZF method, and the digital precoding matrix is:

[0195]

[0196] where \(\mathbf{d}_{i}\) b,i represents the precoding vector of the \(i\) - th user or data stream in the digital precoding matrix, \(\mathbf{H}_{eq}\) represents the equivalent channel matrix between the base station and the user obtained based on channel estimation, and \(\mathbf{H}\) represents the actual multi - user channel matrix.

[0197] The received signal at the ground is:

[0198]

[0199] Meanwhile, the ground access point receiver applies the SIC algorithm to decode the user signals in the order from the worst to the best channel conditions. The user path loss of the RF link is:

[0200]

[0201] where, and represent the satellite transmitter and ground access point receiving antenna gains of the RF link respectively, is the reference distance in the far - field of the antenna, is the length of the RF link, and \(v\) RF is the radio frequency path loss exponent. The SNR of the received signal is calculated by gradually eliminating interference, and the decoding order is arranged according to the RF link channel loss \(h_{i}\) RF in ascending order.

[0202] The present invention is designed for the three - segment hybrid link of high - speed mobile satellite - high - altitude platform - ground access point, and has good application adaptability and deployment conditions. The ground - high - altitude platform link adopts the FSO / RF hybrid mode, and realizes link enhancement through the UAV relay nodes deployed at an altitude between 1 - 2 km, ensuring to avoid the masking effect caused by low - altitude obstacles and improving the link availability. For the FSO link part, the deployment end adopts a channel environment simulation platform based on the geometric loss statistical model of pointing error, atmospheric attenuation, and turbulence modeling, and combines the DDT module to perform Doppler frequency shift compensation processing. In the RF link part, a traffic dynamic clustering module is deployed, and the service types are divided through the EM algorithm combined with the Gaussian mixture model, and a hybrid precoding module (analog + digital precoding) and a receiver - side SIC decoder are integrated. The overall system deployment can complete the software and hardware deployment at the ground access point and the high - altitude platform, realize automatic selection and dynamic scheduling, and support the later algorithm update and optimization.

[0203] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A FSO / RF hybrid communication method based on Doppler compensation and interference suppression, characterized in that The method includes the following steps: Step S1: Construct the satellite-ground transmission link topology; adopt an FSO / RF hybrid link for the link between the high-altitude platform and the ground access point, and construct a topology structure based on the FSO / RF hybrid link; Step S2: Establish an FSO link loss model; the FSO link loss model comprehensively considers atmospheric attenuation, atmospheric turbulence, and AoA fluctuation; two atmospheric turbulence modes, LN and GG, are considered for atmospheric turbulence; deduce the link loss caused by atmospheric attenuation, the channel loss of weak turbulence under the LN model, the channel loss of medium-strong turbulence under the GG model, and the channel loss caused by AoA fluctuation; Step S3: Establish a statistical model for pointing error geometric loss; The statistical model for pointing error geometric loss comprehensively considers the position deviation of the high-altitude platform receiver, the position deviation of the satellite transmitter, and the change in the satellite transmitter tilt angle caused by the high-speed movement of the satellite, and deduces the pointing error; comprehensively consider the effects of atmospheric attenuation, atmospheric turbulence, pointing error, and AoA fluctuation to obtain the channel loss coefficient of the FSO link; Step S4: Establish an RF link loss model and introduce the SR model to calculate the RF link channel loss; Step S5: FSO link Doppler frequency shift compensation; adopt binary phase shift keying modulation and erbium-doped fiber amplifier gain control at the satellite transmitter; adopt the DDT algorithm at the ground access point receiver to compensate for the Doppler frequency shift caused by the high-speed movement of the satellite; Step S6: RF link multi-user interference compensation; the satellite transmitter performs traffic clustering on the accessed services through the EM algorithm, and combines analog precoding and digital precoding for joint signal preprocessing; adopt the SIC algorithm at the ground access point receiver to further suppress multi-user interference.

2. The FSO / RF hybrid communication method based on Doppler compensation and interference suppression according to claim 1, wherein In Step S2, the link loss caused by atmospheric attenuation is expressed as: h al = exp(-Lξ) Among them, h al represents the link loss caused by atmospheric attenuation, ξ represents the attenuation coefficient, and L represents the optical link length; The channel loss of weak turbulence under the LN model is expressed as: Among them, represents the channel loss of weak turbulence under the LN model, is the Rytov variance; Among them, represents the channel loss of moderately strong turbulence under the GG model, the parameters α and β respectively represent the effective numbers of large-scale and small-scale vortices, and Γ(·) is the Gamma function; The channel loss caused by AoA fluctuation is expressed as: where h aoa represents the channel loss caused by the AoA fluctuation, and θ FoV represents the field of view angle of the high-altitude platform receiver, and σ angle,u represents the joint variance of the horizontal and vertical directions of the high-altitude platform.

3. The FSO / RF hybrid communication method based on Doppler compensation and interference suppression according to claim 2, characterized in that In Step S3, the pointing error is expressed as: where h pe represents the pointing error, A represents the fraction of the harvested energy, and η a 2 represents the ratio of the square of the equivalent beam width to the total displacement variance.

4. The FSO / RF hybrid communication method based on Doppler compensation and interference suppression according to claim 3, characterized in that In Step S4, the RF link channel loss is expressed as: where P is the power of the RF link, N0 is the RF link noise power, and γ RF represents the SNR of the RF link, represents the average SNR of the RF link, P L represents the average power of the LOS component, 2b0 represents the average power of the scattered component, 1F1(·, ·, ·) is the confluent hypergeometric function of the first kind, and m is the shape parameter of the Nakagami-m distribution.

5. The FSO / RF hybrid communication method based on Doppler compensation and interference suppression according to claim 4, wherein In Step S5, The output signal of the satellite transmitter is: where b n is a signal symbol, and p(t - nT) is a pulse shaping function. T is the symbol duration, that is, the time interval of each symbol. t - nT represents the signal within the nth symbol time interval, and n represents the symbol time interval index; At the ground access point receiver, the signal after Doppler compensation is: Among them, represents performing a DDT transform on a link with power p n , ω'(t) represents the signal perturbation term, j represents the imaginary unit, and G(l a ) represents the amplification gain, and Ω(τ) is the Doppler frequency shift related to the delay.

Images