A large-scale MIMO wireless channel modeling method for the short-wave frequency band

Abstract

The present invention discloses a method for modeling a large-scale MIMO wireless channel in the short-wave band. The specific steps are as follows: 1) Calculate the distance ranges of three short-wave communication propagation modes, and determine the short-wave communication propagation mode in the current environment; 2) Conduct frequency prediction to obtain the current optimal operating frequency; 3) Conduct large-scale channel parameter modeling; 4) Conduct small-scale channel parameter modeling; 5) Conduct space-time-frequency non-stationarity modeling; 6) Analyze the channel statistical characteristics and verify the model. Due to the limited short-wave communication spectrum resources and the severe ionospheric channel conditions, the system rate is relatively low. Most of the existing short-wave channel models are point-to-point small-scale MIMO, and the system rate gain is not obvious. The present invention establishes a method for modeling a large-scale MIMO wireless channel in the short-wave band, which can support communication scenarios with large-scale MIMO antenna configurations and is compatible with three short-wave communication propagation modes. The wireless channel model based on the present invention has guiding significance for the construction and performance analysis of short-wave large-scale MIMO wireless communication systems.

Description

Technical Field

[0001] The present invention relates to a method for modeling a large-scale MIMO wireless channel in the short-wave band, belonging to the technical field of wireless communication channel modeling. Background Art

[0002] In the International Telecommunication Union (ITU), short waves refer to electromagnetic waves with frequencies in the range of 3 - 30 MHz. Short-wave communication, also known as high-frequency (HF) communication. Short-wave communication has many remarkable advantages. Compared with communication methods such as satellites, microwaves, coaxial cables, and optical fibers, short waves can achieve long-distance communication without the need to establish relay stations. A large number of antennas are used on the base station side of large-scale MIMO, effectively improving the sum rate and reliability of the communication system. As a key technology for the fifth-generation (5G) communication system, large-scale MIMO has been widely applied in frequency bands such as sub-6 GHz and millimeter waves.

[0003] Due to the difficulty in realizing large-scale antenna arrays in the short-wave band, the application of large-scale MIMO in the short-wave band is still scarce. Traditional short-wave channel models are only applicable to the case of a small number of antennas, and the establishment of a large-scale MIMO short-wave channel model urgently needs to be solved. In addition, traditional short-wave channel models can only be used for single propagation modes such as skywave or ground wave, and there is an urgent need for a short-wave channel model that supports multiple propagation modes. Summary of the Invention

[0004] The present invention provides a method for modeling a large-scale MIMO wireless channel in the short-wave band. This modeling method takes into account the near-field spherical wave effect and the spatio-temporal-frequency non-stationary characteristics of the large-scale MIMO channel, and uses a frequency prediction method to accurately set the short-wave operating frequency. At the same time, it supports three short-wave propagation modes: long-distance skywave, near-vertical incident skywave (NVIS), and ground wave. The correctness of the model is verified through channel characteristics or system characteristics such as spatio-temporal-frequency correlation functions, path loss, channel capacity, and singular value spread.

[0005] To achieve the above object, the specific technical solutions of the present invention are as follows:

[0006] A method for modeling a large-scale MIMO wireless channel in the short-wave band, comprising the following steps:

[0007] Step S1, calculate the distance ranges of three short-wave communication propagation modes, and determine the short-wave communication propagation mode in the current environment.

[0008] Step S2: Perform frequency prediction to obtain the current optimal operating frequency.

[0009] Step S3: Conduct large-scale channel parameter modeling.

[0010] Step S4: Conduct small-scale channel parameter modeling.

[0011] Step S5: Perform space-time-frequency non-stationarity modeling.

[0012] Step S6: Analyze channel statistical characteristics and verify the model.

[0013] Furthermore, step S1 specifically includes the following steps:

[0014] Step S101: Calculate the maximum wave propagation distance. The ground wave is divided into two parts, the direct wave and the ground surface wave. The maximum propagation distance of the direct wave is obtained by calculation as where h T and h R are the antenna heights (km) of the transmitter and receiver respectively. And the maximum propagation distance (km) of the ground surface wave is where f is the short-wave operating frequency (MHz). Therefore, the maximum propagation distance (km) of the ground wave is

[0015] Step S102: Calculate the minimum skywave propagation distance. The minimum skywave propagation distance (km) can be expressed as where, β = Z m / (R E +h I ), θ0 is the incident angle of the electromagnetic wave, N max is the maximum ionization concentration of the ionosphere, Z m is the height (km) from the bottom of the ionosphere to the position of the maximum electron concentration, the radius of the earth (km) is R E , and the height of the ionosphere (km) is h I .

[0016] Step S103: Determine the short-wave communication mode. The area between the two distances is the blind spot of short-wave communication. To eliminate this blind spot, the NVIS mode is introduced. D is the distance (km) between the transmitter and the receiver. The division of the three modes is as follows:

[0017] 1) Propagate in the ground wave mode, usually less than 500 km;

[0018] 2) Propagate in the NVIS mode, usually between 500 km and 3000 km;

[0019] 3) Propagated in the skywave mode, it is usually more than 3000 km under normal circumstances.

[0020] Furthermore, the step S2 specifically includes the following steps:

[0021] Step S201: The basic maximum usable frequency (MUF) of each layer of the ionosphere can be obtained through calculation. The basic MUF of the E layer of the ionosphere E = M E f E , M E = 3.94 + 2.80ω - 1.70ω 2 - 0.60ω 3 + 0.96ω 4 , r is the total length of the propagation path (km), f E is the critical frequency (MHz) of the E layer at the midpoint of the path. The basic MUF of the F1 layer of the ionosphere F1 = M F1 f F1 , M F1 = K0 - 0.01(K0 - K 100 )R 12 . K0 = 0.16 + 2.64×10 -3 r - 0.40×10 -6 r 2 , K 100 = -0.52 + 2.69×10 -3 r - 0.39×10 -6 r 2 , f F1 is the critical frequency (MHz) of the F1 layer at the midpoint of the path. The basic C r = 0.74 - 0.591Z - 0.424Z 2 - 0.090Z 3 + 0.088Z 4 + 0.181Z 5 + 0.096Z 6 , Z = 1 - 2r n / r max , r n = r / n0, where n0 is the minimum number of hops for propagation in the F2 layer of the ionosphere, C 3000 is the value of C r at r = 3000 km, f F2 is the critical frequency (MHz) of the F2 layer at the midpoint of the path, f His the magnetic rotation frequency (MHz) at the midpoint of the path. In summary, the total operating MUF of the ionosphere = max{MUF E , MUF F1 , MUF F2 R op}, where R op is the ratio of the MUF of the F2 layer to the basic MUF.

[0022] Step S202: Through the above calculations, the basic MUF of each layer of the ionosphere is obtained. Therefore, the optimal working frequency (OWF) is OWF = max{OWF E , OWF F1 , OWF F2}, OWF E = 0.95MUF E , OWF F1 = 0.95MUF F1 , OWF F2 = R op MUF F2 F l , where F l is the MUF-OWF conversion factor.

[0023] Furthermore, step S3 specifically includes the following steps:

[0024] Step S301: Calculate the shadow fading parameter. The value of the shadow fading parameter follows a lognormal distribution. The median of the distribution can be estimated by a deterministic method, while the standard deviation usually has to be obtained through a large number of measurements under typical conditions. The standard deviation value is between 3 - 10 dB.

[0025] Step S302: Calculate the path loss parameter. The path loss of the sky wave (including the NVIS mode) includes three parts PL S = L f + L i + L g . Among them, the basic transmission loss L f = 32.45 + 20lgf + 20lgr, where r is the total length of the propagation path (km). The ionospheric absorption loss L i = e-∫αdr, where e is the natural constant, and the attenuation coefficient σ is the conductivity, and ε r is the dielectric constant. The ground reflection loss R v is the vertical ground reflection coefficient, and R h is the horizontal ground reflection coefficient. The path loss of the ground wave is PL G= 69.55 + 26.16lgf - 13.82lgh te -α(h re ) + [44.9 - 6.55lgh te lgr - K, h te (h re ) is the effective height (km) of the transmitting (receiving) antenna, and K is an environment-related correction factor.

[0026] Furthermore, step S4 specifically includes the following steps:

[0027] Step S401: The entire shortwave channel matrix H can be expressed as the superposition of three-mode channel matrices: H = where H G , H N , and H S represent the channel matrices in the ground wave, NVIS, and sky wave modes respectively. The function ε(·) is the unit step function. Further substituting the large-scale parameters, we can obtain where P ∈ {G, N, S}.

[0028] Step S402: The short-term fading impulse response of the shortwave channel can be expressed as

[0029]

[0030] where and represent the vertical and horizontal polarization antenna patterns of the p-th transmitting antenna and the q-th receiving antenna. obeys the uniform distribution on (0, 2π]. represents the cross-polarization power ratio. and are the power and delay of the m-th path between the p-th transmitting antenna and the q-th receiving antenna in the n-th scatterer cluster at time t respectively. The carrier frequency f c takes the value of OWF and is obtained through step S202.

[0031] Furthermore, step S5 specifically includes the following steps:

[0032] Step S501: Generate a new scatterer cluster. Assume that the number of paths M in scatterer cluster n n obeys the Poisson distribution is the mean and variance of M n . The virtual delay can be expressed as r τ is the scalar delay, σ τ is the randomly generated delay spread, u nis a uniformly distributed random variable on (0, 1). r τ and σ τ can be obtained through channel measurements. The power of the scatterer clusters can be assumed to have a one-sided exponential delay power spectrum where Z n follows a Gaussian distribution. After the powers of all N scatterer clusters are generated, the normalized power can be expressed as In addition, the von Mises distribution is used to model the azimuth and elevation angles of the channel model. Therefore, the average power of the m n th path can be modeled as is the shadow fading factor that follows a normal distribution. The normalized average power is

[0033] Step S502: The scatterer clusters evolve separately on the array, time, and frequency axes. Assume that the generation rate and recombination rate of the scatterer clusters are λ G and λ R , then the survival probability of the transmit (receive) end clusters is where is the environment correlation factor on the array axis. Therefore, the spatio-temporal-frequency scatterer cluster evolution process is described as follows:

[0034] 1) In the first step, first generate the initial scatterer clusters at time t, as well as randomly generated parameters, such as the number of paths, delay, power, angle parameters, and virtual path delay.

[0035] 2) In the second step, at time t + Δt, two situations may occur, and we discuss them separately. When t is divisible by Δt, first perform the evolution of the clusters on the time axis. The average relative velocity and are respectively characterized as Then the survival probability Pr(Δt) of the clusters after Δt is denoted as where P F is the percentage of mobile clusters in the total number of clusters, is the environmental factor of spatial correlation on the time axis. The evolution process of the clusters on the frequency axis is similar to that on the time axis. The survival probability Pr(Δf) after Δf is where the environmental factor of spatial correlation on the frequency axis, and F(Δf) can both be obtained from channel measurements.

[0036] 3) In the third step, when t is not divisible by Δt, only update the remaining clusters. The first process is to update the geometric relationship of the scatterer clusters from t to t + Δt. First, update the distance vector to Then, update the time delay to through perform first-order filtering to obtain where Υ is a random variable and independent of but has the same distribution as and is a scenario-related parameter that keeps the virtual path consistent.

[0037] Furthermore, the specific steps of step S6 are as follows:

[0038] Step S601: Through the large-scale MIMO wireless channel impulse response, channel statistical characteristic functions including the spatio-temporal-frequency correlation function channel capacity etc. can be derived, laying a foundation for further subsequent analysis.

[0039] Step S602: By comparing with the simulation model and theory, use the spatio-temporal-frequency correlation function, channel capacity and other channel statistical characteristics to verify the large-scale MIMO short-wave wireless channel model respectively. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a schematic flow chart of the large-scale MIMO wireless channel modeling method in the short-wave frequency band in Embodiment 1 of the present invention;

[0041] Figure 2 is a schematic diagram of the large-scale MIMO short-wave wireless communication system in Embodiment 1 of the present invention;

[0042] Figure 3 is a schematic diagram of the large-scale MIMO short-wave wireless channel model in Embodiment 1 of the present invention;

[0043] Figure 4 is a simulation diagram of the channel spatial cross-correlation function in Embodiment 1 of the present invention;

[0044] Figure 5 is a simulation diagram of the channel time auto-correlation function in Embodiment 1 of the present invention;

[0045] Figure 6 is a simulation diagram of the channel capacity in Embodiment 1 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] In order to better understand the purpose, structure and function of the present invention, the following further describes in detail a large-scale MIMO wireless channel modeling method in the short-wave frequency band of the present invention with reference to the drawings.

[0047] Embodiment 1

[0048] See Figure 1 , a large-scale MIMO wireless channel modeling method in the short-wave frequency band provided by this embodiment. The schematic diagram of the system model is shown in Figure 2 , and the schematic diagram of the channel model is shown in Figure 3 . The wireless channel modeling method specifically includes the following steps:

[0049] Step S1: Calculate the distance ranges of three short-wave communication propagation modes, and determine the short-wave communication propagation mode in the current environment.

[0050] Specifically, in this embodiment, step S1 includes the following steps:

[0051] Step S101: Calculate the maximum wave propagation distance. The ground wave is divided into two parts, the direct wave and the ground surface wave. The maximum propagation distance of the direct wave is obtained by calculation as where h T and h R are the antenna heights (km) of the transmitter and the receiver respectively. And the maximum propagation distance (km) of the ground surface wave is where f is the short-wave operating frequency (MHz). Therefore, the maximum propagation distance (km) of the ground wave is

[0052] Step S102: Calculate the minimum skywave propagation distance. The minimum skywave propagation distance (km) can be expressed as where, β = Z m / (R E +h I ), θ0 is the incident angle of the electromagnetic wave, N max is the maximum ionization concentration of the ionosphere, Z m is the height (km) from the bottom of the ionosphere to the position of the maximum electron concentration, the radius of the earth (km) is R E , and the height of the ionosphere (km) is h I .

[0053] Step S103: Short-wave communication mode determination. The area between the two distances is the blind spot of short-wave communication. To eliminate this blind spot, the NVIS mode is introduced. D is the distance (km) between the transmitter and the receiver, and the division of the three modes is as follows:

[0054] 1) Propagate in the ground wave mode, usually less than 500 km;

[0055] 2) Propagate in the NVIS mode, usually between 500 km and 3000 km;

[0056] 3) Propagated in the skywave mode, usually greater than 3000 km under normal circumstances.

[0057] In this embodiment, the transceiver distance can be set to D = 5000 km, and it is determined that the current shortwave working mode is skywave through calculation.

[0058] Step S2: Perform frequency prediction to obtain the current optimal operating frequency.

[0059] Specifically, in this embodiment, step S2 includes the following steps:

[0060] Step S201: The basic maximum usable frequency (MUF) of each layer of the ionosphere can be obtained through calculation. The basic MUF of the E layer of the ionosphere E = M E f E , M E = 3.94 + 2.80ω - 1.70ω 2 - 0.60ω 3 + 0.96ω 4 , r is the total propagation path length (km), and f E is the critical frequency (MHz) of the E layer at the midpoint of the path. The basic MUF of the F1 layer of the ionosphere F1 = M F1 f F1 , M F1 = K0 - 0.01(K0 - K 100 )R 12 , K0 = 0.16 + 2.64×10 -3 r - 0.40×10 -6 r 2 , K 100 = -0.52 + 2.69×10 -3 r - 0.39×10 -6 r 2 , f F1 is the critical frequency (MHz) of the F1 layer at the midpoint of the path. The basic C r = 0.74 - 0.591Z - 0.424Z 2 - 0.090Z 3 + 0.088Z 4 + 0.181Z 5 + 0.096Z 6 , Z = 1 - 2r n / r max , r n= r / n0, where n0 is the minimum number of hops for propagation in the F2 layer of the ionosphere, and C 3000 is the value of C r at r = 3000 km, and f F2 is the critical frequency (MHz) of the F2 layer at the midpoint of the path, and f H is the magnetic rotation frequency (MHz) at the midpoint of the path. In summary, the total working MUF of the ionosphere = max{MUF E , MUF F1 , MUF F2 R op}, where R op is the ratio of the working MUF to the basic MUF of the F2 layer.

[0061] Step S202: Through the above calculations, the basic MUF of each layer of the ionosphere is obtained. Therefore, the Optimal working frequency (OWF) is OWF = max{OWF E , OWF F1 , OWF F2}, OWF E = 0.95MUF E , OWF F1 = 0.95MUF F1 , OWF F2 = R op MUF F2 F l , where F l is the MUF - OWF conversion factor.

[0062] Step S3: Perform large - scale channel parameter modeling.

[0063] Specifically, in this embodiment, Step S3 includes the following steps:

[0064] Step S301: Calculate the shadow fading parameter. The value of the shadow fading parameter follows a log - normal distribution. The median of the distribution can be estimated by a deterministic method, while the standard deviation usually has to be obtained through a large number of measurements under typical conditions. The standard deviation value is between 3 - 10 dB.

[0065] Step S302: Calculate the path loss parameter. The path loss of the sky wave (including the NVIS mode) includes three parts PL S = L f + L i + L g . Among them, the basic transmission loss L f = 32.45 + 20lgf + 20lgr, where r is the total length of the propagation path (km). The ionospheric absorption loss L i= e-∫α dr, where e is the natural constant and α is the attenuation coefficient σ is the conductivity and ε r is the permittivity. The ground reflection loss R v is the vertical ground reflection coefficient, and R h is the horizontal ground reflection coefficient.

[0066] Step S4: Perform small-scale channel parameter modeling.

[0067] Specifically, in this embodiment, Step S4 includes the following steps:

[0068] Step S401: The entire short-wave channel matrix H can be expressed as the superposition of three-mode channel matrices: where h G , H N , and H S represent the channel matrices in the ground wave, NVIS, and sky wave modes respectively. The function ε(·) is the unit step function. Further substituting the large-scale parameters, we can obtain where P ∈ {G, N, S}. Since the working mode in this embodiment is the sky wave, H = H S .

[0069] Step S402: The small-scale fading impulse response of the short-wave channel (sky wave working mode) can be expressed as

[0070]

[0071] where and represent the vertical and horizontal polarization antenna patterns of the p-th transmitting antenna and the q-th receiving antenna. and follow the uniform distribution on (0, 2π]. represents the cross-polarization power ratio. and are the power and delay of the m-th path between the p-th transmitting antenna and the q-th receiving antenna in the n-th scatterer cluster at time t respectively. The carrier frequency takes the value of OWF, which is obtained through Step S202.

[0072] Step S5: Perform space-time-frequency non-stationarity modeling.

[0073] Specifically, in this embodiment, Step S5 includes the following steps:

[0074] Step S501: Generate new scatterer clusters. Assume that the number of paths M in scatterer cluster n n follows the Poisson distribution is M n The mean and variance of. The virtual delay can be expressed as r τ is the scalar delay, σ τ is the randomly generated delay spread, u n is a uniformly distributed random variable on (0,1). r τ and σ τ can be obtained through channel measurements. The power of the scatterer clusters can be assumed to be a one-sided exponential delay power spectrum where, Z n follows a Gaussian distribution. After the powers of all N scatterer clusters are generated, the normalized power can be expressed as In addition, the von Mises distribution is used to model the azimuth and elevation angles of the channel model. Therefore, the average power of the m n th path can be modeled as is the shadow fading factor that follows a normal distribution. The normalized average power is

[0075] Step S502, the scatterer clusters evolve on the array, time, and frequency axes respectively. Assume that the generation rate and recombination rate of the scatterer clusters are λ G and λ R , then the survival probability of the transmitting (receiving) end clusters is where,[[]] is the environment-related factor on the array axis. Therefore, the spatio-temporal-frequency scatterer cluster evolution process is described as follows:

[0076] 1) In the first step, first generate the initial scatterer clusters at time t, as well as randomly generated parameters, such as the number of paths, delay, power, angle parameters, and virtual path delay.[[]]

[0077] 2) In the second step, at time t+Δt, two situations may occur, and we discuss them separately. When t is divisible by Δt, first perform the evolution of the clusters on the time axis. The average relative velocity and are respectively characterized as Then the survival probability Pr(Δt) of the clusters after Δt is denoted as where, P F is the percentage of mobile clusters in the total number of clusters,[[]] is the environmental factor of spatial correlation on the time axis. The evolution process of the clusters on the frequency axis is similar to that on the time axis. The survival probability Pr(Δf) after Δf is where,[[]] The environmental factor of spatial correlation on the frequency axis, and F(Δf) can both be obtained from channel measurements.

[0078] 3) In the third step, when t is not divisible by Δt, only the remaining clusters need to be updated. The first process is to update the geometric relationship of the scatterer clusters from t to t + Δt. First, update the distance vector to Then, update the time delay to Through Perform first-order filtering to obtain where Υ is a random variable and independent of But is identical in distribution to and is a scenario-related parameter that keeps the virtual paths consistent.

[0079] Step S6: Channel statistical characteristic analysis and model verification.

[0080] Specifically, in this embodiment, step S6 includes the following steps:

[0081] Step S601: Through the large-scale MIMO wireless channel impulse response, channel statistical characteristic functions including the spatio-temporal-frequency correlation function Channel capacity etc. can be deduced, laying a foundation for further analysis later.

[0082] Step S602: Refer to Figures 4-6 and use the spatio-temporal-frequency correlation function, channel capacity and other channel statistical characteristics to verify the large-scale MIMO short-wave wireless channel model by comparing with the simulation model and theory respectively.

[0083] In summary, the present invention establishes a large-scale MIMO wireless channel modeling method in the short-wave band, which can support communication scenarios with large-scale MIMO antenna configurations and is compatible with three short-wave communication propagation modes. The wireless channel model based on the present invention has guiding significance for the construction and performance analysis of short-wave large-scale MIMO wireless communication systems.

[0084] It can be understood that the present invention is described by some embodiments. Those skilled in the art know that without departing from the spirit and scope of the present invention, various changes or equivalent replacements can be made to these features and embodiments. In addition, under the teaching of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application belong to the scope protected by the present invention.

Claims

1. A large-scale MIMO wireless channel modeling method in the short-wave frequency band, characterized in that: It includes the following steps: Step S1: Calculate the distance ranges of three shortwave communication propagation modes, and determine the shortwave communication propagation mode in the current environment; Step S2: Conduct frequency prediction to obtain the current optimal operating frequency; Step S3: Conduct large-scale channel parameter modeling; Step S4: Conduct small-scale channel parameter modeling; Step S5: Conduct space-time-frequency non-stationarity modeling; Step S6: Analyze the channel statistical characteristics and verify the model; The specific steps of the said Step S1 include the following steps: Step S101, calculate the maximum wave propagation distance; The ground wave is divided into two parts, the direct wave and the surface wave; By calculation, the maximum propagation distance of the direct wave is where h T and h R are the antenna heights of the transmitting end and the receiving end respectively; And the maximum propagation distance of the surface wave is where f is the shortwave operating frequency; Therefore, the maximum propagation distance of the ground wave is Step S102, calculate the minimum skywave propagation distance; the minimum skywave propagation distance is expressed as where β = Z m / (R E + h I ), θ0 is the incident angle of the electromagnetic wave, N max is the maximum ionization concentration of the ionosphere, Z m is the height from the bottom of the ionosphere to the position of the maximum electron concentration, the radius of the earth is R E , and the height of the ionosphere is h I ; Step S103: Determine the shortwave communication mode; The area between two distances is the blind area of shortwave communication; In order to eliminate this blind spot, the NVIS mode is introduced; D is the distance between the transmitter and the receiver, and the division of the three modes is as follows: 1) Propagated in the ground wave mode, less than 500 km; 2) Propagated in the NVIS mode, between 500 km and 3000 km; 3) Propagated in the skywave mode, greater than 3000 km; The specific steps of the said Step S2 include the following steps: Step S201, calculate the MUF of each layer of the ionosphere; the basic MUF of the E layer of the ionosphere E = M E f E , M E = 3.94 + 2.80ω - 1.70ω 2 - 0.60ω 3 + 0.96ω 4 , r is the total length of the propagation path, f E is the critical frequency of the E layer at the midpoint of the path; the basic MUF of the F1 layer of the ionosphere F1 = M F1 f F1 , M F1 = K0 - 0.01(K0 - K 100 )R 12 , K0 = 0.16 + 2.64×10 -3 r - 0.40×10 -6 r 2 , K 100 = -0.52 + 2.69×10 -3 r - 0.39×10 -6 r 2 , f F1 is the critical frequency of the F1 layer at the midpoint of the path; the basic C r = 0.74 - 0.591Z - 0.424Z 2 - 0.090Z 3 + 0.088Z 4 + 0.181Z 5 + 0.096Z 6 , Z = 1 - 2r n / r max , r n = r / n0, where n0 is the minimum number of hops for propagation in the F2 layer of the ionosphere, C 3000 is the value of C r at r = 3000 km, f F2 is the critical frequency of the F2 layer at the midpoint of the path, f H is the magnetic rotation frequency at the midpoint of the path; In summary, the total operating MUF of the ionosphere = max{MUF E , MUF F1 , MUF F2 R op}, where R op is the ratio of the operating MUF to the basic MUF of the F2 layer; Step S202. Through the above calculations, the basic MUF of each layer of the ionosphere is obtained. Therefore, the OWF is OWF = max{OWF E , OWF F1 , OWF F2}, where OWF E = 0.95MUF E , OWF F1 = 0.95MUF F1 , OWF F2 = R op MUF F2 F l , where F l is the MUF - OWF conversion factor; The specific steps of the said Step S3 include the following steps: Step S301: Calculate the shadow fading parameter; The value of the shadow fading parameter follows a lognormal distribution; The median of the distribution is estimated by a deterministic method, while the standard deviation must be obtained by conducting a large number of measurements; The standard deviation value is between 3 - 10 dB; Step S302: Calculate the path loss parameter; The path loss of the sky wave includes three parts PL S = L f + L i + L g ; where the basic transmission loss L f = 32.45 + 20lgf + 20lgr, r is the total length of the propagation path; The ionospheric absorption loss L i = e-∫αdr, e is the natural constant, and the attenuation coefficient σ is the conductivity, ε r is the dielectric constant; The ground reflection loss R v is the vertical ground reflection coefficient, R h is the horizontal ground reflection coefficient; The path loss of the ground wave is PL G = 69.55 + 26.16lgf - 13.82lgh te - α(h re ) + [44.9 - 6.55lgh te lgr - K, h te is the effective height of the transmitting antenna, h re is the effective height of the receiving antenna, and K is an environment-related correction factor; The specific steps of the said Step S4 include the following steps: Step S401: The entire short-wave channel matrix H is expressed as the superposition of three-mode channel matrices: where H G , H N , and H S represent the channel matrices in the ground-wave, NVIS, and sky-wave modes respectively; the function ε(·) is the unit step function; further substituting the large-scale parameters, we get where P ∈ {G, N, S}; Step S402: The small-scale fading impulse response of the shortwave channel is expressed as Among them, and represent the vertical and horizontal polarization antenna patterns of the p-th transmitting antenna and the q-th receiving antenna; and follow a uniform distribution on (0, 2π]; represents the cross-polarization power ratio; and are respectively the power and delay at time t of the m-th path between the p-th transmitting antenna and the q-th receiving antenna in the n-th scatterer cluster; the carrier frequency f c takes the value of OWF and is obtained through step S202; The specific steps of the said Step S5 include the following steps: Step S501: Generate a new scatterer cluster; the number of paths M in the preset scatterer cluster n n obeys the Poisson distribution is M n The mean and variance of; the virtual time delay is expressed as r τ is the scalar time delay, σ τ is the randomly generated delay spread, u n is a uniform distribution random variable on (0,1); r τ and σ τ are obtained through channel measurement; the power of the scatterer cluster is preset to a one-sided exponential delay power spectrum where, Z n obeys the Gaussian distribution; after the powers of all N scatterer clusters are generated, the normalized power is expressed as In addition, the azimuth angle and elevation angle of the channel model are modeled using the von Mises distribution; thus, the m n th path average power is modeled as is the shadow fading factor obeying the normal distribution; the normalized average power is Step S502, the scatterer clusters evolve respectively along the array, time, and frequency axes; the generation rate and recombination rate of the preset scatterer clusters are λ G and λ R , then the survival probability of the transmitting and receiving end clusters is wherein, is the environment-related factor on the array axis; therefore, the evolution process of the space-time-frequency scatterer clusters is described as follows: 1) In the first step, first generate the initial scatterer cluster at time t, as well as randomly generated parameters, including the number of paths, time delay, power, angle parameters, and virtual path time delay; 2) Second step, at time t+Δt, two situations may occur: when t is divisible by Δt, first, the evolution of the clusters on the time axis is carried out; the average relative velocity and are respectively characterized as Then the survival probability Pr(Δt) of the clusters after Δt is denoted as where P F is the percentage of mobile clusters in the total number of clusters, is the environmental factor of spatial correlation on the time axis; the evolution process of the clusters on the frequency axis is the same as that on the time axis; the survival probability Pr(Δf) after Δf is where the environmental factor of spatial correlation on the frequency axis, and F(Δf) can both be obtained from channel measurements; 3) In the third step, when t is not divisible by Δt, only update the remaining clusters; the first process is to update the geometric relationship of the scatterer clusters from t to t + Δt; first, update the distance vector to Then, update the time delay to Through Perform first-order filtering to obtain where Υ is a random variable and independent of But is the same distribution as and is a scene-related parameter that makes the virtual path consistent; The specific steps of the said Step S6 include the following steps: Step S601: Derive the space-time-frequency correlation function including the large-scale MIMO wireless channel impulse response Channel capacity Channel statistical characteristic function, laying the foundation for further subsequent analysis; Step S602: By comparing with the simulation model and theory, use the channel statistical characteristics of the space-time-frequency correlation function and channel capacity to verify the large-scale MIMO shortwave wireless channel model respectively.

Images