A sea surface scene low-altitude unmanned aerial vehicle millimeter wave communication channel modeling method

Abstract

This invention discloses a method for modeling millimeter-wave communication channels for low-altitude unmanned aerial vehicles (UAVs) in maritime scenarios. First, a three-dimensional geometrical random channel model is constructed, classifying scatterers into three categories: a bispherical surface representing near-field scatterers at both ends of the transmission path, a bifocal ellipsoid describing atmospheric particles in high-humidity salt fog, and partially cylindrical surfaces simulating dynamic sea surface and wave scatterers. Second, based on this model, the channel statistical characteristics are derived, including the channel impulse response formed by line-of-sight, primary, and secondary scattering components, the spatiotemporal correlation function, and the Doppler power spectral density, where the scatterer angles follow a mixed probability distribution. Finally, a path loss model is constructed: based on the two-path model, an additional attenuation factor related to salt fog concentration and atmospheric humidity is introduced, with parameters determined by sea state levels. This invention is well-suited to high-humidity salt fog environments and dynamic sea state characteristics, accurately characterizing the spatiotemporal characteristics of the channel and providing a theoretical basis for link planning and performance evaluation of millimeter-wave communication systems for maritime UAVs.

Description

Technical Field

[0001] This invention belongs to the field of low-altitude millimeter-wave communication channel modeling technology, and particularly relates to a method for modeling millimeter-wave communication channels for low-altitude unmanned aerial vehicles (UAVs) in maritime scenarios. Specifically, it is a method for constructing a low-altitude three-dimensional geometric random scattering statistical channel model that can fully consider the attenuation effect of high humidity salt spray and the dynamic sea surface scattering effect. This model is used to accurately evaluate the time-varying characteristics and transmission attenuation of the channel in low-altitude UAV millimeter-wave communication links under complex sea conditions, thereby laying a theoretical foundation for the design of maritime low-altitude UAV millimeter-wave communication systems. Background Technology

[0002] With the rapid development of the low-altitude economy, the application of unmanned aerial vehicle (UAV) technology in maritime logistics, maritime monitoring, fisheries management, coastal patrol, and maritime emergency rescue is booming. The demand for self-organizing network communication based on UAV platforms is also increasing, and how to design an efficient and reliable UAV communication network for specific complex sea conditions is becoming an urgent problem to be solved.

[0003] In low-altitude scenarios over the sea, multiple physical effects, including sea surface reflection, wave splash scattering, and attenuation from high-humidity salt spray, are significant factors affecting the quality of communication links between UAVs. To accurately characterize the impact of these factors, such as ocean waves, on communication systems, corresponding statistical models are needed. Existing stochastic distribution models such as Rayleigh, Rice, and log-normal distributions are applicable to certain uniform propagation scenarios, but they are insufficient to accurately describe the time-varying scattering and attenuation mechanisms in the sea environment. A more accurate geometric stochastic statistical scattering model is required.

[0004] For signal propagation attenuation, free-space propagation models or two-path models are commonly used. However, due to the reflection effect of the sea surface on electromagnetic wave propagation, the free-space propagation model is not suitable. In contrast, the two-path model combines the direct path and the sea surface reflection path, and can more accurately describe the propagation attenuation of low-altitude millimeter-wave signals over the sea surface. However, the traditional two-path model neither considers the diffuse scattering effect caused by sea surface roughness, nor can it characterize the additional absorption and scattering losses caused by high humidity atmosphere and salt spray particles. Therefore, there is an urgent need for a modeling method that can accurately describe the millimeter-wave communication channel of low-altitude UAVs over the sea surface.

[0005] The paper "Broadband Non-Stationary UAV Air-to-Air Communication Channel Model" uses a Gaussian Markov motion model to model the speed and flight direction of the UAV. This model focuses on analyzing the influence of the UAV's own motion, but does not fully consider the impact of complex environments. The paper "Characterizing UAV Air-to-Air Channel Characteristics with an Extended 3D Ellipsoid Model" provides a precise description of received power in terms of time delay and direction of arrival based on existing 3D ellipsoid models, but does not discuss its applicability to channel characteristics at higher frequency bands such as millimeter waves (e.g., more severe path loss, more significant blocking effects). Chinese patent CN202111233049.8, entitled "A Non-Stationary Geometric Random Channel Modeling Method for Underwater Acoustic Communication," uses two sets of bi-cluster models to describe randomly distributed scatterers on the seabed and sea surface, respectively. However, there are significant differences between underwater acoustic channels and low-altitude channels at the sea surface, and ultrasonic waves have very long wavelengths, resulting in transmission characteristics that differ significantly from millimeter waves. The invention, with publication number CN202211603220.4 and titled "A Geometrically Random and Beam-Domain Wireless Channel Modeling Method for Ultra-Large-Scale MIMO," supports spherical waves and spatial non-stationary characteristics, and can achieve the conversion from geometrically random to beam-domain channel models under near-field conditions. However, this model assumes ideal array structures and antenna elements. In actual ultra-large-scale MIMO systems, hardware impairments can significantly affect beam-domain characteristics, and the model does not discuss the robustness to actual hardware impairments. No relevant literature or patents on low-altitude millimeter-wave communication channel modeling over sea were found.

[0006] In summary, current research on millimeter-wave communication channel models for low-altitude UAVs in maritime scenarios has failed to adequately consider the dynamic sea surface scattering conditions. Therefore, there is an urgent need for a method that can accurately describe the combined propagation effects caused by wave splash, salt spray distribution, and rough sea surface in the low-altitude maritime environment, thereby laying a theoretical foundation for the design of millimeter-wave communication systems for low-altitude UAVs. Summary of the Invention

[0007] The purpose of this invention is to construct a modeling method for millimeter-wave communication channels of low-altitude unmanned aerial vehicles (UAVs) in marine scenarios. This method introduces a composite attenuation factor related to salt spray mass concentration and atmospheric relative humidity, and couples it into the link budget of geometry-based stochastic channel modeling. Simultaneously, it constructs a double sphere, a complete ellipsoid, and a partial cylinder to simulate near-field scatterers, high-humidity salt spray particle scatterers, and dynamic marine surface scatterers, respectively. This allows for a unified framework that simultaneously describes the absorption and scattering attenuation caused by marine scenarios and the multipath scattering effects under complex sea conditions, achieving an accurate expression of the time-varying characteristics and signal propagation attenuation of millimeter-wave communication channels for low-altitude UAVs over the marine surface.

[0008] The technical solution adopted in this invention is as follows:

[0009] A method for modeling millimeter-wave communication channels for low-altitude unmanned aerial vehicles (UAVs) in a sea surface scenario includes the following steps:

[0010] Step 1: Divide the scatterers in the millimeter-wave communication scenario of low-altitude UAVs over the sea into three categories and construct a three-dimensional geometric random scattering statistical channel model. Specifically: For the discrete scatterers of seabirds existing within a set range at both ends of the UAVs, describe them using two spherical models centered on the transmitting and receiving UAVs respectively; For the atmospheric particles of high humidity salt fog suspended in the air under the communication scenario, describe them using a complete ellipsoidal model centered on the transmitting and receiving UAVs; For the dynamic sea surface and splashing waves between the transmitting and receiving ends, describe them using a partial cylindrical model centered on the midpoint between the transmitting and receiving ends.

[0011] Step 2: Based on the three-dimensional geometric random scattering statistical channel model, obtain the statistical characteristics of the channel, specifically including: impulse response function, spatiotemporal correlation function and Doppler power spectral density function; and considering the sea surface mirror effect, superimpose attenuation factors related to salt spray mass concentration and atmospheric relative humidity on the two-path propagation model to construct a path propagation loss model.

[0012] Step 3: Obtain the modeling results of the millimeter-wave communication channel for low-altitude UAVs in the sea surface scene, including the models and functions described in Step 1 and Step 2.

[0013] Furthermore, in step 1, the axial height of some cylinders is greater than the semi-major axis length of the complete ellipsoid; the radius of some cylinders is dynamically related to the sea surface roughness parameter.

[0014] Furthermore, the impulse response function H(t) of the channel is N T ×N R 3D matrix, N T and N R Let be the number of antenna elements in the uniform linear array configured at the transmitter and receiver, respectively. In the matrix elements of H(t), starting from the Tth digit at the transmitter... p From the Rth antenna element to the receiver p The channel impulse response of each antenna element is represented by h. pq (t) represents p = 1, 2, ..., N T q = 1, 2, ..., N R h pq (t) is composed of the superposition of the line-of-sight propagation component, multiple primary scattering components, and multiple secondary scattering components in the complex impulse response:

[0015]

[0016] In the formula, This represents the impulse response of the line-of-sight (LoS) component; The i-th type of primary scattering SS component is represented by i = 1, 2, 3, 4, which correspond to the scattering body of the transmitter spherical model, the scattering body of the receiver spherical model, the scattering body of the complete ellipsoidal model, and the scattering body of the partial cylindrical model, respectively. and These represent five different types of secondary scattering DS components. The subscript of DS indicates the scattering bodies the signal passes through in sequence. 12 This indicates that the signal transmitted by the transmitter is scattered by the transmitter's near-field scatterer, then by the receiver's near-field scatterer, and finally reaches the receiver. DS 13 This indicates that the signal is scattered by the near-field scatterer of the transmitter, then by atmospheric particle scatterers, and finally reaches the receiver. DS 14 This indicates that the signal is scattered by the near-field scatterer of the transmitter, then by the dynamic sea surface scatterer, and finally reaches the receiver. DS 32 This indicates that the signal is scattered by atmospheric particle scatterers, then by scatterers near the receiver, and finally reaches the receiver. DS 42 This means that the signal is scattered by a scatterer in the dynamic sea surface area, then by a scatterer in the near-field area of ​​the receiver, and finally reaches the receiver.

[0017]

[0018]

[0019] In the formula, λ represents the operating wavelength; K is the Rice factor; and These represent the normalized power coefficients of each non-line-of-sight nLoS component, and these normalized power coefficients satisfy the condition that their sum is 1; and Representing the Los component and SS respectively i Components, DS 12 Components, DS 13 Components, DS 14 Components, DS 32 Components and DS 42 The Doppler frequency shift of the components, i = 1, 2, 3, 4; N1, N2, N3, N4 represent the effective number of scatterers in the transmitter spherical model, receiver spherical model, complete ellipsoidal model, and partial cylindrical model, respectively; ε LoS , and Representing the Los component and SS respectively i Components, DS 12 Components, DS 13 Components, DS 14 Components, DS 32 Components and DS 42 The propagation distance of the component.

[0020] Furthermore, in step 2, the space-time correlation function R pq,p'q' (d T ,d R The values ​​of ,t,Δt) are obtained by weighted summation of the spatiotemporal correlation functions of the LoS component and each nLoS component, where the spatiotemporal correlation functions of the LoS component and each nLoS component are obtained by integrating the probability distribution function of the scattering angle of the corresponding component.

[0021] Furthermore, the probability distribution function of the scattering angle α is a mixed probability distribution, expressed as:

[0022] f(α)=ω spec f spec (α)+ω diff f diff (α)+ω spray f spray (α)

[0023] Among them, the specular reflection component f spec (α) follows a Gaussian distribution; diffuse scattering component f diff (α) follows a Von-Mises distribution; the splash component f spray (α) follows a uniform distribution; ω spec ω diff and ω spray These represent the specular scattering weight, diffuse scattering weight, and splash weight, respectively. The weight coefficients of each scattering component are dynamically related to the sea state level.

[0024] Furthermore, in step 2, the Doppler power spectral density is obtained by performing a Fourier transform on the time autocorrelation function of the spatiotemporal correlation function.

[0025] Furthermore, in step 2, the attenuation factor α total Salt spray absorption attenuation factor α salt With water vapor absorption attenuation factor α vapor Together they constitute the model as follows:

[0026] α total =α salt +α vapor =γ×(MCC) m ×(RH) n

[0027] In the formula, MCC is the salt spray mass concentration; RH is the relative humidity; the composite attenuation coefficient γ is related to the operating frequency; and the concentration index m and humidity index n are the relevant fitting coefficients.

[0028] The advantages of this invention compared to the prior art are:

[0029] 1) This invention proposes a low-altitude hybrid three-dimensional geometric random model, which uses a double sphere, a double-focal ellipsoid, and a partial cylinder to represent near-field scatterers, high-humidity salt spray particle scatterers, and dynamic sea surface scatterers, respectively. Under a unified framework, a hybrid probability distribution model is adopted to integrate three mechanisms: specular reflection, diffuse scattering, and wave splashing, to accurately reproduce the composite propagation scenario of the sea surface.

[0030] 2) This invention introduces a composite attenuation factor that is strongly correlated with frequency and accurately calculates the absorption and scattering effects of salt spray particles in the millimeter wave band based on Mie theory.

[0031] 3) This invention establishes a dynamic mapping system between sea state level and model parameters, where all key parameters are related to sea state level. Attached Figure Description

[0032] Figure 1 This is a flowchart of the millimeter-wave communication channel modeling method for low-altitude UAVs in a sea surface scene according to the present invention.

[0033] Figure 2 This is a schematic diagram of the scatterer and parameters in the channel model proposed in this invention.

[0034] Figure 3 This is the horizontal structure diagram of the channel model proposed in this invention.

[0035] Figure 4 This is a spatial correlation characteristic diagram of an embodiment of the present invention.

[0036] Figure 5 This is a time-related characteristic diagram of an embodiment of the present invention.

[0037] Figure 6 This is a Doppler power spectrum of an embodiment of the present invention.

[0038] Figure 7 This is a schematic diagram of path loss according to an embodiment of the present invention. Detailed Implementation

[0039] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0040] A method for modeling millimeter-wave communication channels for low-altitude UAVs in a sea surface scenario, such as Figure 1 As shown, it includes the following steps:

[0041] Step 1: Construct a millimeter-wave communication channel propagation model for low-altitude UAVs in a sea surface scenario.

[0042] Step 1-1: Constructing a Statistical Model of Scatterers. In the marine environment, scattering near the UAV platform mainly originates from its own structural components and potential seabirds. These scatterers exhibit three-dimensional randomness in spatial distribution. Therefore, this invention uses a bifocal ellipsoid to simulate the near-field scatterers of the UAV platform. In atmospheric regions with high humidity and salt fog, the scattering field formed by suspended particles typically exhibits symmetry about the transmit / receive link axis. Using a bifocal ellipsoid with the transmitting (Tx) and receiving (Rx) UAVs as focal points can effectively describe this symmetrical scattering structure and cover the main distribution range of salt fog particles. The ellipsoid can characterize the reflection and scattering paths of salt fog particles on signals, thereby simulating the resulting multipath effect. By adjusting the semi-major and semi-minor axis dimensions of the ellipsoid, the intensity and angular distribution characteristics of salt fog scattering can be effectively controlled. For scatterers in dynamic sea surface areas, a partial cylinder is used for characterization. Its axial height is greater than the semi-major axis length of the ellipsoid in high humidity and salt fog environments, and the cylinder radius is related to the sea surface wave height to reflect the influence of wave scale on the distribution range of scatterers. Figure 2 This is a schematic diagram of the scatterer and parameters in the channel model proposed in this invention. The basic geometric elements in the channel model and their definitions are shown in Table 1.

[0043] Table 1. Basic geometric parameters and definitions of the channel model

[0044]

[0045]

[0046] Steps 1-2: Constructing the signal propagation model for each component. During signal propagation, the signal reaching the receiver will experience significant attenuation after multiple reflections and scatterings by scatterers such as the dynamic sea surface, waves, and salt spray particles. The constructed geometric scattering model primarily considers primary and secondary scattering at line-of-sight and non-line-of-sight distances, with each propagation component independent of the others. Primary scattering includes reflections and scatterings from the sea surface, waves, and salt spray particles. and The first scattering propagates. Secondary scattering involves successive scattering via... and Secondary scattering, sequentially through and Secondary scattering, sequentially through and Secondary scattering, sequentially through and Secondary scattering, sequentially through and The signal undergoes secondary scattering. The parameters and definitions of each signal component are shown in Table 2.

[0047] Table 2. Parameters and definitions for each signal component.

[0048]

[0049] Step 2: The communication transmitter (Tx) and receiver (Rx) are each equipped with N T and N R For a uniform linear array antenna with N elements, the impulse response H(t) of the channel can be expressed as N. T ×N R 3D matrix representation.

[0050] Step 2-1: Derive the path distance from Tx to Rx or the scatterer using geometric modeling. ε LoS =ε p,q This represents the propagation path distance of the Loss component; These represent the single scattering components SS and SS, respectively. i (i = 1, 2, 3, 4) corresponds to the path T p →S(n1)→R q T p →S(n2)→R q T p →S(n3)→R q T p →S(n4)→R q The propagation distance; They represent the secondary scattering components DS, respectively. 12 DS 13 DS 14 DS 32 DS 42 via path T p →S(n1)→S(n2)→R q T p →S(n1)→S(n3)→R q T p →S(n1)→S(n4)→R q T p →S(n3)→S(n2)→R q T p →S(n4)→S(n2)→R q The propagation distance; and Representing the Los component and SS respectively i (i = 1, 2, 3, 4) components, DS 12 Components, DS 13 Components, DS 14 Components, DS 32 Components and DS 42 The Doppler frequency shift of the component; and Let represent the normalized power coefficients of each non-line-of-sight (nLoS) component, and let these normalized power coefficients sum to 1.

[0051] Step 2-2: Based on the Ricean fading channel, the channel impulse response from Tx to Rx is obtained as follows:

[0052]

[0053] Among them, the complex impulse response of the LoS component Represented as:

[0054]

[0055] SS i Complex impulse response of (i = 1, 2, 3, 4) components Represented as:

[0056]

[0057] DS 12 The component part is represented as:

[0058]

[0059] DS 13 The component part is represented as:

[0060]

[0061] DS 14 The component part is represented as:

[0062]

[0063] DS 32 The component part is represented as:

[0064]

[0065] DS 42 The component part is represented as:

[0066]

[0067] Where K is the Rice factor and λ represents the wavelength of the carrier wave;

[0068] Step 3: Considering the non-stationary characteristics of the channel, both Tx and Rx are in motion, and the relative motion between the transmitting and receiving ends causes Doppler frequency shift.

[0069] Doppler shift of LoS component It is generated by the relative motion between Tx and Rx, that is

[0070]

[0071] In the formula, f T =v T / λ、f R =v R / λ represent the maximum Doppler frequency shift caused by the motions Tx and Rx, respectively. and The horizontal departure angle (AAoD) and horizontal arrival angle (AAoA) of the Loss component are represented. and This represents the vertical departure angle (EAoD) and vertical arrival angle (EAoA) of the Loss component.

[0072] In nLoS multipath components, SS i (i = 1, 2, 3, 4) components, DS 12 Components, DS 13 Components, DS 14 Components, DS 32 Components, DS 42 The Doppler frequency shifts of the components are expressed as follows:

[0073]

[0074]

[0075] in and Indicates S(n) i The horizontal and vertical angles of arrival of the scattered radio waves describe the signal's arrival at S(n) from the center of the Tx antenna array, in the horizontal and vertical directions, respectively. i The angle of S(n). Similarly, the signal leaves S(n) i The horizontal departure angle to the center of the Rx antenna array is defined as follows: and vertical departure angle Similarly, via S(n) i The horizontal and vertical departure angles of the scattered signal are expressed as follows: and The signal leaves S(n) i The horizontal and vertical departure angles corresponding to the receiving end are defined as follows: and

[0076] Step 4: Further determine the propagation distance of each component of the channel.

[0077] Step 4-1, as follows Figure 3 As shown, the geometric relationship between the horizontal scattering angles can be obtained by simplifying the horizontal structure diagram of the model:

[0078]

[0079] We can obtain,

[0080]

[0081] Similarly,

[0082]

[0083] By the Law of Cosines:

[0084]

[0085]

[0086] In the vertical direction, it satisfies

[0087]

[0088] From the above derivation, the conversion relationship between the corresponding scattering angles can be obtained.

[0089] Step 4-2: In this model, the azimuth and elevation angles of the same scattering component are independent. Electromagnetic wave scattering in a sea surface environment is a complex process involving multiple mechanisms. It mainly includes three scattering components: specular reflection component, diffuse scattering component, and splash component. This invention uses a mixed probability distribution model to describe the scattering body angle α, which is:

[0090] f(α)=ω spec f spec (α)+ω diff f diff (α)+ω spray f spray (α)

[0091] Among them, the specular reflection component f spec (α) is contributed by the large-scale wavefront of the sea surface, and its scattering angle distribution is concentrated near the specular reflection direction, following a Gaussian distribution. μ spec σ represents the direction of specular reflection. spec Standard deviation; diffuse scattering component f diff (α) is contributed by small-scale ripples on the sea surface, and its scattering direction is influenced by the dominant wave direction, following a Von-Mises distribution. μ diff The mean direction is indicated by the location of the average direction, k represents the degree of concentration of the data around the mean direction, and I0(k) is the 0th-order modified Bessel function of the first kind; the splash component f spray(α) Caused by spray and bubbles generated by wave breaking, the scattering angle distribution is assumed to be isotropic and follows a uniform distribution.

[0092] The weighting coefficient ω of each scattering component spec ω diff and ω spray It is dynamically related to sea state levels, as detailed in Table 3.

[0093] Table 3 Weight Value Comparison Table

[0094]

[0095]

[0096] Step 5, the Space-Time Correlation Function (STCF) is defined as follows:

[0097]

[0098] Where E[·] represents the expectation function, (·) * This represents the complex conjugate operation, and Δt represents the time interval.

[0099] Step 5-1: The spatiotemporal correlation function of the specific LoS component and the primary scattering components SS1, SS2, and SS3 is expressed as follows:

[0100]

[0101] This invention uses a partially cylindrical shape to describe a dynamic scatterer in a sea surface scene. On a vast sea surface, the scattering environment can be considered omnidirectional on the horizontal plane. Based on the analysis of typical wave splash heights in this scene, the range of values ​​for the pitch angle is set as follows: Therefore, the STCF of the SS4 component is represented as:

[0102]

[0103] DS 12 The STCF representation of the component is as follows:

[0104]

[0105] DS 13 The STCF representation of the component is as follows:

[0106]

[0107] DS 14 The STCF representation of the component is as follows:

[0108]

[0109] DS 32 The STCF representation of the component is as follows:

[0110]

[0111] DS 42 The STCF representation of the component is as follows:

[0112]

[0113] Step 5-2: Under certain conditions, the time correlation function (ACF) and spatial correlation function (CCF) of the channel can be derived from the channel STCF. Set the antenna element spacing to 0, i.e., d... T =d R =0, the channel's ACF can be obtained from the channel STCF. Similarly, by setting Δt=0, the CCF can be obtained from the channel STCF.

[0114] Performing a Fourier transform on the ACF yields the channel Doppler power spectral density, which is:

[0115]

[0116] Step 6: Taking into account the attenuation of signal propagation by high-humidity salt fog particles in low-altitude sea scenarios, construct a path propagation loss model based on the two-path propagation model.

[0117] Step 6-1: Constructing the basic propagation loss model. This invention considers the reflection effect of the sea surface in real-world scenarios, using the Loss-of-Stake (LoS) path between Tx and Rx and the reflection path from the sea surface as the main propagation paths to construct a signal propagation loss model.

[0118]

[0119] Among them, G T With G R Here, λ represents the transmit antenna gain and the receive antenna gain, respectively; λ is the signal operating wavelength; Γ is the sea surface reflection coefficient; M and N are distance parameters. Among them, H T With H R Tx and Rx represent the antenna heights, respectively, and D represents the propagation distance of the LoS path.

[0120] Step 6-2: In a marine environment, high-humidity salt spray particles absorb and scatter electromagnetic waves. When analyzing wave propagation attenuation in a discrete random medium, an attenuation coefficient is typically used. The attenuation coefficient reflects the attenuation effect of particles on electromagnetic waves. To quantify the absorption and scattering effects of the marine environment on electromagnetic waves, this invention introduces a characteristic attenuation α related to the salt spray mass concentration and atmospheric relative humidity. total Represented as:

[0121] α total =α salt +α vapor =γ×(MCC) m ×(RH) n

[0122] Wherein, MCC is the salt spray mass concentration (g / m³). 3 ), where RH is relative humidity, the composite attenuation coefficient γ is related to the operating frequency, and the concentration index m and humidity index n are the relevant fitting coefficients.

[0123] Sea state levels are divided into five grades: E0-E4, based on Dow wave magnitude and sea surface wind speed. The characteristic attenuation α is calculated. total Under sea state E0, wind speeds are low, the sea surface is calm, wave splashing is weak, the concentration of salt mist in the air is extremely low, and humidity is moderate. At this time, the additional attenuation of millimeter wave propagation by salt mist particles is negligible. Under sea state E1, ripples and small waves appear on the sea surface, a small amount of sea foam droplets begin to form, the salt mist concentration increases slightly, humidity increases, and the initial hygroscopic growth of salt mist particles becomes apparent, producing observable scattering and absorption effects on electromagnetic waves. Under sea state E2, whitecap waves appear, wave splashing is significant, the salt mist concentration further increases, humidity continues to rise, and the hygroscopic effect of particles increases. Stronger winds and larger particle sizes cause significant Mie scattering and absorption attenuation of millimeter waves. Under sea state E3, the wind and waves are larger, and the number of breaking waves on the sea surface increases, stirring up a large number of water droplets and salt spray particles, reaching a high concentration. The humidity is close to saturation, and attenuation is mainly dominated by high-concentration, large-particle-size deliquescent salt spray particles, resulting in significantly enhanced signal attenuation. Under sea state E4, the sea surface is turbulent, with spray and foam filling the air. The salt spray concentration is extremely high, the humidity is saturated, and the hygroscopic effect of particles reaches its maximum. Both scattering and absorption attenuation are extremely strong, seriously affecting the communication link. The system obtains marine environmental parameters such as current wind speed and significant wave height in real time through the environmental perception module, and determines the current sea state level and its corresponding composite attenuation factor parameter set according to the preset sea state level classification strategy (as shown in Table 4).

[0124] Table 4. Sea State Level Comparison Table

[0125]

[0126] Step 6-3: Combining the basic propagation loss with the propagation loss caused by low-altitude high-humidity salt spray particles, the propagation loss of a signal with a working wavelength of λ after propagating along a path of distance D is:

[0127]

[0128] The invention will now be described in further detail with reference to examples.

[0129] The basic parameters of the embodiments selected in this invention specification are as follows:

[0130] D = 1km, R T =15m, R R =15m, f c =35GHz, v T =20m / s, v R =20m / s, H T =200m, H R =180m, antenna element spacing d at the transmitting and receiving ends T =d R =λ / 2 (λ is the operating wavelength), α vT =α vR =0, 2a=1200m, b=100m, h=1400m, R=150m, K=0.5.

[0131] After implementation, at sea state E2, the spatial autocorrelation function and temporal autocorrelation function for k values ​​of 0.5, 2, 5, 10, and 15 are as follows: Figure 4 , Figure 5 As shown. When k=5, the Doppler power spectrum function results are as follows. Figure 6 As shown. Path loss is as follows. Figure 7 As shown.

Claims

1. A method for modeling millimeter-wave communication channels for low-altitude unmanned aerial vehicles (UAVs) in a sea surface scenario, characterized in that, Includes the following steps: Step 1: Divide the scatterers in the millimeter-wave communication scenario of low-altitude UAVs over the sea into three categories and construct a three-dimensional geometric random scattering statistical channel model. Specifically: For the discrete scatterers of seabirds existing within a set range at both ends of the UAVs, describe them using two spherical models centered on the transmitting and receiving UAVs respectively; For the atmospheric particles of high humidity salt fog suspended in the air under the communication scenario, describe them using a complete ellipsoidal model centered on the transmitting and receiving UAVs; For the dynamic sea surface and splashing waves between the transmitting and receiving ends, describe them using a partial cylindrical model centered on the midpoint between the transmitting and receiving ends. Step 2: Based on the three-dimensional geometric random scattering statistical channel model, obtain the statistical characteristics of the channel, specifically including: impulse response function, spatiotemporal correlation function and Doppler power spectral density function; and considering the sea surface mirror effect, superimpose attenuation factors related to salt spray mass concentration and atmospheric relative humidity on the two-path propagation model to construct a path propagation loss model. Step 3: Obtain the modeling results of the millimeter-wave communication channel for low-altitude UAVs in the sea surface scene, including the models and functions described in Step 1 and Step 2; In step 1, the axial height of some cylinders is greater than the semi-major axis of the complete ellipsoid; the radius of some cylinders is dynamically related to the sea surface roughness parameter. In step 2, the impulse response function of the channel... for 3D matrix and The number of antenna elements in the uniform linear array configured for the transmitting and receiving ends, respectively. In the matrix elements, starting from the transmitter... The antenna array element to the receiver The channel impulse response of each antenna element is used express, , , It is composed of the superposition of the line-of-sight propagation component, multiple primary scattering components, and multiple secondary scattering components: In the formula, Indicates line of sight distance The impulse response of the component; Indicates the first Types of primary scattering SS components, These correspond to the scatterer model of the transmitting end sphere, the scatterer model of the receiving end sphere, the scatterer model of the complete ellipsoid, and the scatterer model of the partial cylinder, respectively. , , , and These represent five different types of secondary scattering DS components. The subscript of DS indicates the scattering bodies the signal passes through in sequence. 12 This indicates that the signal transmitted by the transmitter is scattered by the transmitter's near-field scatterer, then by the receiver's near-field scatterer, and finally reaches the receiver. DS 13 This indicates that the signal is scattered by the near-field scatterer of the transmitter, then by atmospheric particle scatterers, and finally reaches the receiver. DS 14 This indicates that the signal is scattered by a near-field scatterer from the transmitter, then by a dynamic sea surface scatterer, and finally reaches the receiver. DS 32 This indicates that the signal is scattered by atmospheric particle scatterers, then by scatterers near the receiver, and finally reaches the receiver. DS 42 This means that the signal is scattered by a scatterer in the dynamic sea surface area, then by a scatterer in the near-field area of ​​the receiver, and finally reaches the receiver. In the formula, Indicates the operating wavelength; Rice factor; , , , , and Representing each non-line-of-sight distance The normalized power coefficients of the components, which sum to 1; , , , , , and They represent Quantity, Quantity, Quantity, Quantity, Quantity, Components and The Doppler frequency shift of the components, i=1,2,3,4; , , , These represent the effective number of scatterers in the transmitter spherical model, receiver spherical model, complete ellipsoidal model, and partial cylindrical model, respectively. , , , , , and They represent Quantity, Quantity, Quantity, Quantity, Quantity, Components and The propagation distance of the component; In step 2, the space-time correlation function Depend on Components and each The weighted summation of the space-time correlation functions of the components is obtained, where Components and each The spatiotemporal correlation function of the component is obtained by integrating the probability distribution function of the scattering angle of the corresponding component; In step 2, the Doppler power spectral density function is obtained by performing a Fourier transform on the time autocorrelation function of the spatiotemporal correlation function.

2. The method for modeling millimeter-wave communication channels of low-altitude UAVs in a sea surface scene according to claim 1, characterized in that, The probability distribution function of the scattering angle α is a mixed probability distribution, expressed as: Among them, the specular reflection component Follows a Gaussian distribution; diffuse scattering component Follows a Von-Mises distribution; splash component It follows a uniform distribution; , and These represent the specular scattering weight, diffuse scattering weight, and splash weight, respectively. The weight coefficients of each scattering component are dynamically related to the sea state level.

3. The method for modeling millimeter-wave communication channels of low-altitude UAVs in a sea surface scene according to claim 1, characterized in that, Attenuation factor in step 2 Salt spray absorption attenuation factor With water vapor absorption attenuation factor Together they constitute the model as follows: In the formula, This refers to the salt spray mass concentration. Relative humidity; composite attenuation coefficient Related to operating frequency; concentration index and humidity index These are the relevant fitting coefficients.

Images