METHOD FOR DESIGNING TIME-DOMAIN NON-STATIONARY V2V MIMO COMMUNICATION CHANNEL EMULATOR

Abstract

A method for designing a time-domain non-stationary V2V MIMO communication channel emulator includes determining basic parameters for the V2V MIMO communication channel; generating a V2V 2D time-domain non-stationary communication channel environment, by using a MATLAB, that is, the numbers of the scatterers and the positions of the scatterers and the like; importing parameters generated in the previous step into a hardware simulation platform to calculate communication channel parameters for clusters, such as an angle distribution and a power distribution, writing a Verilog code for running, and eventually calculating to obtain a channel impulse response of the time-domain non-stationary V2V MIMO communication channel; and comparing with a statistical characteristic of a theoretical communication channel model, and designing an appropriate hardware diagram of a communication channel emulator. The method supports the simulation of time-domain non-stationary V2V MIMO communication channel, filling the gap in the field of communication channel emulators.

Description

TECHNICAL FIELD

The present disclosure relates to the technical field of channel modeling and channel emulators, and especially relates to a method for designing an a time-domain non-stationary vehicle-to-vehicle (V2V) multiple-input multiple-output (MIMO) communication channel emulator.

BACKGROUND

Before a new communication system be commercialized, the new communication system needs to be practically tested in a corresponding environment. Although the system can be tested on-site in different locations, this method has a relative high cost and is easy to be influenced by surrounding environment. In addition, propagation conditions of a real external field are difficult to be reproduced by comparative analyses of simulation results. A more practical method is to create suitable and stable simulation environments for channels, and then evaluate performances of the communication system in these environments. The channel environments can be controllably and repeatably simulated by channel emulators, which is used for a consistency testing, a performance testing, and an interoperability testing of communication systems. This means that there is no need for an on-site testing, and efficiencies of time and cost can be greatly improved. At present, there is no emulator for a time-domain non-stationary V2V communication channel that supports a birth-death process of clusters and considers macro cells, micro cells, and micro-micro cells.

In summary, it is necessary to establish a method for designing a time-domain non-stationary V2V MIMO communication channel emulator.

SUMMARY

In view of this, the objectives of the present disclosure are to provide a method for designing the time-domain non-stationary V2V MIMO communication channel emulator, so as to accurately and stably test a V2V MIMO communication system.

In order to achieve above objectives, the technical solutions adopted by the present disclosure are as follows.

A method for designing a time-domain non-stationary V2V MIMO communication channel emulator is provided. The method includes the following steps.

In Step S1, the basic parameters for the V2V MIMO communication channel are determined.

In Step S2, a V2V two-dimensional (2D) time-domain non-stationary communication channel environment is generated by using a MATLAB, and the environment specifically includes the number of scatterers and positions of the scatterers, a random phase of a non-line-of-sight (NLoS) path, an angle spread of the NLoS path, a sine function lookup table, and an arctangent function lookup table.

In Step S3, parameters generated in Step S2 are imported into a hardware simulation platform to calculate communication channel parameters for the clusters, and the communication channel parameters include an angle distribution and an amplitude distribution, a Verilog code is written for running, and a channel impulse response of the time-domain non-stationary V2V MIMO communication channel is eventually obtained by calculation.

In Step S4, a comparison is performed with a statistical characteristic of a theoretical communication channel model, and an appropriate hardware diagram of a communication channel emulator is designed.

Preferably, in Step S1, in a geometry-based stochastic model for the time-domain non-stationary V2V MIMO communication channel, the basic parameters for the V2V MIMO communication channel include the number of a simulation time point, a simulation time interval, a position of a transmitter, a position of a receiver, a velocity of the transmitter, a velocity of the receiver, a Rice factor, an angle spread coefficient, a total link power, a line-of-sight (LoS) path, a power ratio of a single-bounce (SB) path, a power ratio of a double-bounce (DB) path, the number of initial scatterers, a velocity of the initial scatterers, the number of sub-paths in each cluster, a generation rate for scatterers, a disappearance rate for the scatterers, a motion ratio of the scatterers, a relative coordinate of a receiver antenna and a relative coordinate of a transmitter antenna.

Preferably, Step S2 specifically includes as follows.

In Step S201, the numbers of the scatterers and the positions of the scatterers are generated, and an average survival probability P_remain of effective scatterers within a time period Δt is expressed as

$P_{remain}=\exp \left\{-{\gamma }_R\left[P_m\left({\overrightarrow{v}}_{S_1}+{\overrightarrow{v}}_{S_2}\right)\Delta t+{\overrightarrow{v}}_{Rx}-{\overrightarrow{v}}_{Tx}\Delta t\right]\right\},$

where γR denotes a generation rate of the scatterers, P_m denotes a motion percentage, {right arrow over (v)}_s1 and {right arrow over (v)}_s2 denote an average velocity of scatterer S₁ and an average velocity of scatterer S₂, respectively. {right arrow over (v)}_Tx and {right arrow over (v)}_Rx denote a velocity of the transmitter and a velocity of the receiver, respectively.

A probability function P(N_new(Δt)=k) of the number of newly generated scatterers is expressed as

$P\left(N_{new}\left(\Delta t\right)=k\right)=\frac{{\lambda }^k}{k!}{e}^{-\lambda },k=0,1,\dots ,\mathrm{and}$ $\lambda =\frac{{\gamma }_G}{{\gamma }_R}\left(1-P_{remain}\right)$

where λ denotes a parameter for a Poisson distribution probability distribution function; and

• • γ_G denotes a disappearance rate of the scatterers, k denotes a parameter; then, an average number of the scatterers E{N(t)} is:

$E\left\{N\left(t\right)\right\}=\frac{{\gamma }_G}{{\gamma }_R}.$

Eventually, the numbers of the scatterers within each time period Δt are obtained by calculation, which are generated through a MATLAB, and are sequentially stored in a file with a coe suffix.

Position coordinates of the scatterers are uniformly and randomly generated in a rectangular area around the receiver and the transmitter, which are generated through the MATLAB, and are sequentially stored in the file with a coe suffix.

In Step S202, the random phase of a NLoS path is generated, and the random phase of a NLoS path follows a uniform distribution in [−π, π), which are generated through the MATLAB, and are sequentially stored in the file with a coe suffix.

In Step S203, the angle spread of the NLoS path is generated, whose calculation formula is expressed as

${\phi }_{AoD}^m\left(t\right)=AS\left({\theta }_{AoD}\right)Y_{AoD}^m,$ ${\phi }_{AoA}^m\left(t\right)=AS\left({\theta }_{AoA}\right)Y_{AoA}^m,$

where φ_AoD^m(t) and φ_AoA^m(t)denote an angle spread of an angle of departure (AoD) and an angle spread of an angle of arrival (AoA), respectively, AS(θ_AoD)and AS(θ_AoA)denote an angle spread coefficient for the AoD and an angle spread coefficient for the AoA, respectively, and Y_AoD^m and Y_AoA^m follow a standard normal distribution N(0,1), respectively. The angle spread of the AD and the angle spread of the AoA are generated through the MATLAB, and are sequentially stored in the file with a coe suffix.

In Step S204, the sine function lookup table is generated, and 65536 points are uniformly sampled within one cycle to be symmetric about a y-axis, and function values are amplified to 4096 times, which are generated through the MATLAB, and are sequentially stored in the file with a coe suffix.

In Step S205, the arctangent function lookup table is generated, and 524288 points are uniformly sampled around an origin point with an interval 1/65536 between each sampling point to be symmetric about the y-axis, and function values are amplified to 2¹⁵/π times, which are generated through the MATLAB, and are sequentially stored in the file with a coe suffix.

Preferably, in Step S3, the angle distributions of the clusters are expressed as θ_AoD^LoS(t) θ_AoA^LoS(t), θ_AoD^i,n(t) and θ_AoA^i,n(t), the amplitude distribution is expressed as H_qp^LoS(t), H_qp^SBi(t) and H_qp^DB(t); an angle of the clusters is determined by the coordinate of the transmitter, the coordinate of the receiver, and the coordinates of the scatterers; the amplitude is determined by the Rice factor, the total power, a proportion of a SB ray and a DB ray to a total scattering power on the NLoS path, the number of an i-th class scatterers at a time instant t, and the number of sub-paths in each cluster of the NLoS path, and Step S3 specifically includes the following steps.

In Step S301, the AD of all paths and the AoA of all paths are generated, whose calculation formulas are expressed as

${\theta }_{AoD}^{LoS}\left(t\right)=\arctan \frac{R{x}^y\left(t\right)-T{x}^y\left(t\right)}{R{x}^x\left(t\right)-T{x}^x\left(t\right)},$ ${\theta }_{AoA}^{LoS}\left(t\right)=\arctan \frac{T{x}^y\left(t\right)-R{x}^y\left(t\right)}{T{x}^x\left(t\right)-R{x}^x\left(t\right)},$ ${\theta }_{\mathrm{AoD}}^{i,n}\left(t\right)=\arctan \frac{S_i^y\left(t\right)-R{x}^y\left(t\right)}{S_{i,n}^x\left(t\right)-R{x}^x\left(t\right)},$ ${\theta }_{AoA}^{i,n}\left(t\right)=\arctan \frac{S_i^y\left(t\right)-R{x}^y\left(t\right)}{S_{i,n}^x\left(t\right)-R{x}^x\left(t\right)},$

where S_i,n^x(t), Tx^x(t) and Rx^x(t) denote a horizontal coordinate of the n-th scatterer in the i-th class (i=1,2,3), a horizontal coordinate of the transmitter, and a horizontal coordinate of the receiver, respectively, and S_i,ny(t), Tx^y(t)and Rx^y(t)denote a vertical coordinate of the n-th scatterer in the i-th class, a vertical coordinate of the transmitter, and a vertical coordinate of the receiver, respectively.

In Step S302, a sub-path angle of the NLoS path is generated, whose calculation formulas are expressed as

${\theta }_{AoD}^{i,n,m}\left(t\right)={\phi }_{AoD}^m\left(t\right)+{\theta }_{\mathrm{AoD}}^{i,n},$ ${\theta }_{\mathrm{AoA}}^{i,n,m}\left(t\right)={\phi }_{AoA}^m\left(t\right)+{\theta }_{\mathrm{AoA}}^{i,n},$

where θ_AoD^i,n,m(t)and θ_AoA^i,n,m(t)denote an AoD of the m-th sub-path and an AoA of the m-th sub-path, respectively, φ_AoD^m(t)and φ_AoA^m(t) denote an angle spread value for the AoD of the m-th sub-path and an angle spread value for the AoA of the m-th sub-path, respectively, and φ_AoD^i,n and φ_AoA^i,n denote an average value for the AoDs of clusters passing through the n-th scatterer in the i-th class and an average value for the AoAs of the clusters passing through the n-th scatterer in the i-th class, respectively.

In Step S303, a time delay value is generated, whose calculation formula is expressed as

${\tau }^{LoS}\left(t\right)=\frac{\sqrt{{\left[R{x}^x\left(t\right)-T{x}^x\left(t\right)\right]}^2+{\left[R{x}^y\left(t\right)-T{x}^y\left(t\right)\right]}^2}}{c}$ ${\tau }_n^{{\mathrm{SB}}_i}\left(t\right)=\frac{\sqrt{\begin{array}{c}{\left[S_{i,n}^x\left(t\right)-{\mathrm{Tx}}^x\left(t\right)\right]}^2+{\left[S_{i,n}^y\left(t\right)-{\mathrm{Tx}}^y\left(t\right)\right]}^2+\\ {\left[S_{i,n}^x\left(t\right)-{\mathrm{Rx}}^x\left(t\right)\right]}^2+{\left[S_{i,n}^y\left(t\right)-{\mathrm{Rx}}^y\left(t\right)\right]}^2\end{array}}}{c}$ ${\tau }_{n_1,n_2}^{\mathrm{DB}}\left(t\right)=\frac{\sqrt{\begin{array}{c}{\left[S_{1,n}^x\left(t\right)-{\mathrm{Tx}}^x\left(t\right)\right]}^2+{\left[S_{1,n}^y\left(t\right)-{\mathrm{Tx}}^y\left(t\right)\right]}^2+{\left[S_{1,n}^x\left(t\right)-S_{2,n}^x\left(t\right)\right]}^2+\\ {\left[S_{1,n}^y\left(t\right)-S_{2,n}^y\left(t\right)\right]}^2+{\left[S_{2,n}^x\left(t\right)-{\mathrm{Rx}}^x\left(t\right)\right]}^2+{\left[S_{2,n}^y\left(t\right)-{\mathrm{Rx}}^y\left(t\right)\right]}^2\end{array}}}{c}$

where, τ^LoS(t)denotes a time delay value for a LoS path, τ_n^SBi(t) denotes a time delay value for a SB path cluster passing through the n-th scatterer in the i-th class, τ_n1,n2^DB(t)denotes a time delay value for a DB path cluster passing through the n₁-th scatterer in the first class and the n₂-th scatterer in the second class, and c denotes a velocity of light.

In Step S304, Doppler phase values ϕ^LoS(t), ϕ_n,m^SBi(t) and ϕ_n1,n2m^DB(t) are generated, whose calculation formulas are expressed as

$f^{LoS}\left(t\right)=f_c\left(\frac{{\overrightarrow{v}}_{Tx}^T{\hat{r}}_{Tx}^{LoS}\left(t\right)}{c}+\frac{{\overrightarrow{v}}_{Rx}^T{\hat{r}}_{Rx}^{LoS}\left(t\right)}{c}\right),$

where f^LoS(t) denotes a Doppler frequency of the LoS path, f_c denotes a carrier frequency; and {circumflex over (r)}_TX^LOS(t)=(cosθ_AoD^LoS(t), sinθ_AoD^LoS(t))

${\hat{r}}_{Rx}^{LoS}\left(t\right)=\left(\cos {\theta }_{AoA}^{Los}\left(t\right),\sin {\theta }_{AoA}^{Los}\left(t\right)\right)$ ${\varphi }^{LoS}\left(t\right)={\int }_{t_0}^{t}2\pi f^{LoS}\left(t^{\prime }\right)d{t}^{\prime },$

where {circumflex over (r)}_TX^LoS(t) and θ_RX^LoS(t) denote a unit vector for the AoD of the LoS path and a unit vector for the AoA of the LoS path, respectively, and φ^LoS(t) denotes a Doppler phase of the LoS path; and

$f_{n,m}^{S{B}_i}\left(t\right)=f_c\left(\frac{{\overrightarrow{v}}_{n_i,\mathrm{Tx}}^T{\hat{r}}_{\mathrm{Tx},n,m}^{\mathrm{SB}}\left(t\right)}{c}+\frac{{\overrightarrow{v}}_{n_i,\mathrm{Rx}}^T{\hat{r}}_{\mathrm{Rx},n,m}^{\mathrm{SB}}\left(t\right)}{c}\right),$

f_n,m^SBi(t) denotes a Doppler frequency of the m-th sub-path of the SB path passing through the n-th scatterer in the i-th class, and {right arrow over (v)}_ni,Tx={right arrow over (v)}_Tx−{right arrow over (v)}_ni

${\overrightarrow{v}}_{n_i,Rx}={\overrightarrow{v}}_{Rx}-{\overrightarrow{v}}_{n_i}$ ${\hat{r}}_{\mathrm{Tx},n,m}^{S{B}_i}\left(t\right)=\left(\cos {\theta }_{AoD}^{i,n,m}\left(t\right),\sin {\theta }_{AoD}^{i,n,m}\left(t\right)\right)$ ${\hat{r}}_{\mathrm{Rx},n,m}^{S{B}_i}\left(t\right)=\left(\cos {\theta }_{\mathrm{AoA}}^{i,n,m}\left(t\right),\sin {\theta }_{\mathrm{AoA}}^{i,n,m}\left(t\right)\right)$ ${\varphi }_{n,m}^{S{B}_i}\left(t\right)={\int }_{t_0}^{t}2\pi f_{n,m}^{S{B}_i}\left(t^{\prime }\right){\mathrm{dt}}^{\prime },$

where {right arrow over (v)}_ni,Tx and {right arrow over (v)}_ni,Rx denote a relative velocity of the transmitter with the n-th scatterer and a relative velocity of the receiver with the n-th scatterer, respectively; {circumflex over (r)}_Tx,n,m^SBi(t) and {circumflex over (r)}_Rx,n,m^SBi(t) denote a unit vector for an AoD of the m-th sub-path of the SB path passing through the n-th scatterer in the i-th class, and a unit vector for an AoA of the m-th sub-path of the SB path passing through the n-th scatterer in the i-th class, respectively, and φ_n,m^SBi(t) denotes a Doppler phase of the m-th sub-path of the SB path passing through the n-th scatterer in the i-th class;

$f_{n_1,n_2,m}^{\mathrm{DB}}\left(t\right)=f_c\left(\frac{{\overrightarrow{v}}_{n_1,\mathrm{Tx}}^T{\widehat{r}}_{\mathrm{Tx},n_{1}m}^{\mathrm{DB}}\left(t\right)}{c}+\frac{{\overrightarrow{v}}_{n_2,\mathrm{Rx}}^T{\widehat{r}}_{\mathrm{Rx},n_{2}m}^{\mathrm{DB}}\left(t\right)}{c}\right),$

where f_n1,n2m^DB(t) denotes a Doppler frequency of the moth sub-path of the DB path passing through the n₁-th scatterer in the first class and the n₂-th scatterer in the second class; and

${\overrightarrow{v}}_{n_1,\mathrm{Tx}}={\overrightarrow{v}}_{\mathrm{Tx}}-{\overrightarrow{v}}_{n_1}$ ${\overrightarrow{v}}_{n_2,\mathrm{Rx}}={\overrightarrow{v}}_{\mathrm{Rx}}-{\overrightarrow{v}}_{n_2}$ ${\widehat{r}}_{\mathrm{Tx},n_1,m}^{\mathrm{DB}}\left(t\right)=\left(\cos {\theta }_{\mathrm{AoD}}^{1,n_1,m}\left(t\right),\sin {\theta }_{\mathrm{AoD}}^{1,n_1,m}\left(t\right)\right)$ ${\widehat{r}}_{\mathrm{Rx},n_2,m}^{\mathrm{DB}}\left(t\right)=\left(\cos {\theta }_{\mathrm{AoA}}^{2,n_2,m}\left(t\right),\sin {\theta }_{\mathrm{AoA}}^{2,n_2,m}\left(t\right)\right)$ ${\varphi }_{n_1,n_2,m}^{\mathrm{DB}}\left(t\right)={\int }_{t_0}^{t}2{\mathrm{\pi f}}_{n_1,n_2,m}^{\mathrm{DB}}\left(t^{\prime }\right){\mathrm{dt}}^{\prime },$

where {circumflex over (r)}_TX,n1,m^DB(t), and {circumflex over (r)}_RX,n2,m^DB(t) denote a unit vector for an AoD of the m-th sub-path of a cluster of the DB path passing through the n₁-th scatterer in the first class and the n₂-th scatterer in the second class, and a unit vector for an AoA of the m-th sub-path of the cluster of the DB path passing through the n₁-th scatterer in the first class and the n₂-th scatterer in the second class, respectively, and ϕ_n1,n2m^DB(t) denotes a Doppler phase of the m-th sub-path of the cluster of the DB path cluster passing through the n₁-th scatterer in the first class and the n₂-th scatterer in the second class.

In Step S305, an antenna phase value is generated, whose calculation formula is expressed as follows.

A relative coordinate {right arrow over (d)}_Tx of a transmitter antenna is:

${\overrightarrow{d}}_{\mathrm{Tx}}={\left[d_{\mathrm{Tx}}^x,d_{\mathrm{Tx}}^y\right]}^T,$

where d_Tx^x denotes a horizontal ordinate of {right arrow over (d)}_Tx and d_Tx^y, denotes a vertical ordinate of {right arrow over (d)}_Tx.

A phase difference ψ_Tx(t) of the transmitter antenna is:

${\psi }_{\mathrm{Tx}}\left(t\right)=2\pi f_c\frac{{\widehat{r}}_{\mathrm{Tx}}^T\left(t\right)·{\overrightarrow{d}}_{\mathrm{Tx}}}{c},$

where {circumflex over (r)}_Tx(t) denotes a unit vector for an AoD at the time instant t.

A relative coordinate {right arrow over (d)}_Rx of the receiver antenna i:

${\overrightarrow{d}}_{\mathrm{Rx}}={\left[d_{\mathrm{Rx}}^x,d_{\mathrm{Rx}}^y\right]}^T,$

where d_Rx^x denotes a horizontal ordinate of {right arrow over (d)}_Rx, d_Rx^y denotes a vertical ordinate of {right arrow over (d)}_Rx.

A phase difference ψ_Rx(t) of the receiver antenna is:

${\psi }_{\mathrm{Rx}}\left(t\right)=2\pi f_c\frac{{\widehat{r}}_{\mathrm{Rx}}^T\left(t\right)·{\overrightarrow{d}}_{\mathrm{Rx}}}{c},$

where {circumflex over (r)}_Rx(t) denotes a unit vector for an AoA at the time instant t.

A total antenna phase value ψ(t) is:

$\psi \left(t\right)={\psi }_{\mathrm{Tx}}\left(t\right)+{\psi }_{\mathrm{Rx}}\left(t\right).$

In Step S306, amplitude values are generated, whose calculation formulas are expressed as:

$H^{\mathrm{LoS}}\left(t\right)=\sqrt{\frac{K_{\mathrm{qp}}{P}_{\mathrm{qp}}}{K_{\mathrm{qp}}+1}}$ $H^{{\mathrm{SB}}_i}\left(t\right)=\sqrt{\frac{{\xi }_{{\mathrm{SB}}_i}{P}_{\mathrm{qp}}}{\left(K_{\mathrm{qp}}+1\right)N_i\left(t\right)M}}$ $H_{\mathrm{qp}}^{\mathrm{DB}}\left(t\right)=\sqrt{\frac{{\xi }_{\mathrm{DB}}{P}_{\mathrm{qp}}}{\left(K_{\mathrm{qp}}+1\right)N_1\left(t\right)N_2\left(t\right)M}},$

where, H_qp^LoS(t) denotes a channel impulse response amplitude value for a LoS path between the q-th receiving antenna and the p-th transmitting antenna, H_qp^SBi(t) denotes a channel impulse response amplitude value for a sub-path of a SB path passing through the scatterer in the i-th class located between the q-th receiving antenna and the p-th transmitting antenna, H_qp^DB(t) denotes a channel impulse response amplitude value for the sub-path of the DB path between the q-th receiving antenna and the p-th transmitting antenna, K_qp denotes a Rice factor of a p-q link, P_qp denotes a total power of the p-q link, ξ_SBi and ξ_SBi denote the proportion of the total scattering power of the SB ray and the DB ray on the NLoS path, respectively, N_i(t)denotes the number of the scatterer in the i-th class at the time instant t, and M denotes the number of the sub-paths in each cluster of the NLoS path.

In Step S307, channel impulse response is generated, whose calculation formula is expressed as:

$h_{\mathrm{qp}}^{\mathrm{LoS}}\left(t,\tau \right)=H^{\mathrm{LoS}}\left(t\right)e^{-j2\pi f_c{\tau }^{\mathrm{LoS}}\left(t\right)}{e}^{j{\varphi }^{\mathrm{Los}}\left(t\right)}{e}^{j{\psi }^{\mathrm{LoS}}\left(t\right)}\delta \left(\tau -{\tau }^{\mathrm{LoS}}\left(t\right)\right)$ $h_{\mathrm{qp}}^{\mathrm{SB}}\left(t,\tau \right)=\sum _{i=1}^3\sum _{n_i=1}^{N_i\left(t\right)}{H}^{{\mathrm{SB}}_i}\left(t\right)e^{-j2\pi f_c{\tau }_n^{{\mathrm{SB}}_i}\left(t\right)}{e}^{j{\phi }_{n,m}^{{\mathrm{SB}}_i}\left(t\right)}{e}^{j{\varphi }_{n,m}^{{\mathrm{SB}}_i}\left(t\right)}{e}^{j{\psi }_{n,m}^{{\mathrm{SB}}_i}\left(t\right)}\delta \left(\tau -{\tau }_n^{{\mathrm{SB}}_i}\left(t\right)\right)$ $h_{\mathrm{qp}}^{\mathrm{DB}}\left(t,\tau \right)=\sum _{n_1,n_2=1}^{N_1\left(t\right),N_2\left(t\right)}{H}_{n_1,n_2,m}^{\mathrm{DB}}\left(t\right)e^{-j2\pi f_c{\tau }_{n_1,n_2}^{\mathrm{DB}}}{e}^{j{\phi }_{n_1,n_2,m}^{\mathrm{DB}}\left(t\right)}{e}^{j{\varphi }_{n_1,n_2,m}^{\mathrm{DB}}\left(t\right)}{e}^{j{\psi }_{n_1,n_2,m}^{\mathrm{DB}}\left(t\right)}\delta \left(\tau -{\tau }_{n_1,n_2}^{{\mathrm{DB}}_i}\left(t\right)\right)$ $h_{\mathrm{qp}}\left(t,\tau \right)=h_{\mathrm{qp}}^{\mathrm{LoS}}\left(t,\tau \right)+h_{\mathrm{qp}}^{\mathrm{SB}}\left(t,\tau \right)+h_{\mathrm{qp}}^{\mathrm{DB}}\left(t,\tau \right),$

where h_qp^LoS(t,τ) denotes a channel impulse response for the LoS path between the q-th receiving antenna and the p-th transmitting antenna, h_qp^SB(t,τ) denotes the channel impulse response for the SB path between the q-th receiving antenna and the p-th transmitting antenna, h_qp^DB(t,τ) denotes a channel impulse response for the DB path between the q-th receiving antenna and the p-th transmitting antenna, f_c denotes a carrier center frequency, τ denotes a time delay, ϕ denotes a Doppler phase, ψ denotes an antenna phase difference, φ_n,m^SBi(t) and φ_n1,n2m^DB(t) denote a random phase of the m-th SB sub-path passing through the n-th scatterer in the i-th class and a random phase of the m-th SB sub-path passing through the n₁-th scatterer in the first class and the n₂-th scatterer in the second class, respectively; the channel impulse response is imported from the file with a coe suffix and a calculated channel impulse response is exported to a txt file.

Preferably, in Step S4, formulas for the statistical characteristic specifically include followings:

In Step S401, a calculation formula of a time autocorrelation function TACF curve is expressed as:

$r_{\mathrm{qp},q^{\prime }{p}^{\prime }}=E\left\{{h_{\mathrm{qp}}\left(t,\tau \right)\left[h_{q^{\prime }{p}^{\prime }}\left(t,\tau \right)\right]}^{*}\right\},$

where h_qp(t,τ) denotes a channel impulse response between the q-th receiving antenna and the p-th transmitting antenna in a case where a time period is t and a time delay is τ, h_q′p′(t,τ) denotes a channel impulse response between the q′-th receiving antenna and the p′-th transmitting antenna in a case where a time period is t and a time delay is τ;(·)* denotes a conjugate complex of(·)

In Step S402, a calculation formula of a spatial cross-correlation function(SCCF) curve is expressed as

$r_{\mathrm{qp}}\left(\Delta t\right)=E\left\{{h_{\mathrm{qp}}\left(t,\tau \right)\left[h_{\mathrm{qp}}\left(t+\mathrm{\Delta t},\tau \right)\right]}^{*}\right\}.$

In Step S403, a calculation formula of a delay power spectral density(PSD)curve is expressed as

$\rho \left(\tau ,{\theta }_{\mathrm{AoA}}^{i,n}\right)={h_{\mathrm{qp}}\left(t,\tau \right)}^2{❘}_{{\theta }_{\mathrm{AoA}}^{i,n}}.$

K_qp, ξ_SBi ; and ξ_DB in the model are related to the macro cells, the micro cells, and the micro micro cells, as well as a SB path and a DB path. The model proposed by the present disclosure can adapt to various V2V propagation environments by adjusting the model parameters ξ_SBi , ξ_DB, and the Rice factor K_qp. In the macro cell scenarios, due to a relative large distance between the transmitter and the receiver, the DB rays carry more energy than that of the SB rays(the relative large distance between the transmitter and the receiver leads to a greater independence of an AoD and an AoA, that is, ξ_DB>max{ξ_SB1, ξ_SB2}>>ξ_SB3, and the received signal power mainly comes from the SB rays and the DB rays of the dual loop model. Therefore, the Rice factor K_qp and the energy parameter ξ_SB3 are extremely little, even close to zero. This means that in the macro cell scenario, it can be characterized by using a dual loop model that ignores the LoS radial component. Compared with the macro cell scenario, in the micro cell scenario and the micro-micro cell scenario, vehicular traffic density (VTD)significantly affects the channel characteristics. In consideration of the impacts of VTD on channel statistics, the present disclosure distinguishes the mobile cars and the stationary roadside environments(such as buildings, trees, parked cars, etc.)around the transmitter and the receiver. Therefore, the present disclosure simulates a mobile car by using a dual loop model and describes a stationary roadside environment by using an elliptical model. For low VTD, since the LoS path has a high power, the value for K_qp is large. In addition, the received scattering power mainly comes from the signal reflected by the stationary roadside environment described by the scatterers freely located on the ellipse. Since the mobile car represented by the scatterers located on the dual loop is sparse, it is more likely to be a SB rather than a DB, indicating a validity of ξ_SB3>max{ξ_SB1, ξ_SB2}>ξ_DB. Under a condition of a high VTD, the value for K_qp is smaller than that in a low VTD scenario, and due to the relative large number of the mobile cars, the DB rays of the dual loop model carry more energy than those of the SB rays of the dual loop model and the elliptical model, that is, ξ_DB>max{ξ_SB1, ξ_SB2, ξ_SB3}. Thus, the micro cell scenario and the micro-micro cell scenario in consideration of VTD can be well characterized by utilizing the combined dual loop model and the elliptical model with a LoS radial component.

The beneficial effects of the present disclosure are in the following.

The present disclosure can provide a method for designing a time-domain non-stationary V2V MIMO communication channel emulator, so as to accurately and stably test a V2V MIMO communication system.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 illustrates a flow chart diagram of a method for designing a time-domain non-stationary V2V MIMO communication channel emulator provided in Embodiment 1 of the present disclosure.

FIG. 2 illustrates a schematic diagram of a geometric stochasticity-based model for the time-domain non-stationary V2V MIMO communication channel provided in Embodiment 1 of the present disclosure.

FIG. 3 illustrates a schematic diagram of a hardware block diagram of a time-domain non-stationary V2V MIMO communication channel emulator provided in Embodiment 1 of the present disclosure.

FIG. 4 illustrates a schematic diagram of a TD sum-of-frequency-modulation(SoFM) module in a hardware block diagram of a time-domain non-stationary V2V MIMO communication channel emulator provided in the Embodiment 1 of the present disclosure.

FIG. 5 illustrates a schematic diagram of TACF verification results provided in the Embodiment 1 of the present disclosure, where f_c=500 MHZ, K_qp=1, ξ_SBi =0.2, i=1, 2, 3, ξ_DB=0.4, P_qp=30 dBm, M=8, {right arrow over (v)}_Tx=(10,0), {right arrow over (v)}_Rx=(5,0)and {right arrow over (v)}_S1(2)=(1,0).

FIG. 6 illustrates a schematic diagram of SCCF verification results provided in the Embodiment 1 of the present disclosure, where f_c=500 MHZ, K_qp=1, ξ_SBi =0.2, i=1, 2, 3, ξ_DB=0.4, P_qp=30 dBm, M=8, {right arrow over (v)}_Tx=(10,0), {right arrow over (v)}_Rx=(5,0)and {right arrow over (v)}_S1(2)=(1,0).

FIG. 7 illustrates a schematic diagram of SCCF simulation results provided in the Embodiment 1 of the present disclosure, where f_c=500 MHZ, K_qp=1, ξ_SBi =0.2, i=1, 2, 3, ξ_DB=0.4, P_qp=30 dBm, M=8, {right arrow over (v)}_Tx=(10,0), {right arrow over (v)}_Rx=(5,0)and {right arrow over (v)}_S1(2)=(1,0), γ_G=0.08/m, γ_R=0.03/m, and P_m=0.6.

FIG. 8 illustrates a schematic diagram of simulation results for delayed PSD hardware platform provided in the Embodiment 1 of the present disclosure, where f_c=500 MHZ, K_qp=1, ξ_SBi =0.2, i=1, 2, 3, ξ_DB=0.4, P_qp=30 dBm, M=8, {right arrow over (v)}_Tx=(10,0), {right arrow over (v)}_Rx=(5,0)and {right arrow over (v)}_S1(2)=(1,0), γ_G=0.08/m, γ_R=0.03/m, and P_m=0.6.

DETAILED DESCRIPTION OF THE EMBODIMENTS

In order to clarify the objectives, the technical solutions, and the advantages of the embodiments of the present disclosure to be clearer, the technical solutions in the embodiments of present disclosure will be clearly and completely described in conjunction with the accompanying drawings. Obviously, the described embodiments are one part of the embodiments of the present disclosure, not all of them. Based on the embodiments in the present disclosure, all other embodiments obtained by a person skilled in the art without creative labor fall within the protection scope of the present disclosure.

Embodiment 1

With reference to FIGS. 1 to 8, provided in this embodiment is a method for designing a time-domain non-stationary V2V MIMO communication channel emulator. The method for designing the communication channel emulator specifically includes the following steps.

In Step S1, basic parameters for the V2V MIMO communication channel are determined.

Specifically, in this embodiment, c a method for establishing a geometry-based stochastic channel model (GBSM). Firstly, the application scenario is determined as V2V, and the initial coordinates of the transmitter and the receivers are determined as (8,8) and (50,50) at the same time. The carrier center frequency is set as 500 MHz, the antenna spacing distance between the transmitter and the receiver is set as 1 meter, the elevation angle and the horizontal angle of the transmitting antenna array and the receiving antenna array are set as 0 and 0, the motion velocity of the transmitter and the motion velocity of the receiver are set as 10 m/s and 5 m/s, respectively, and the direction of the motion is that the elevation angle is 0 and the horizontal angle is 0. K_qp=1, ξ_SBi=0.2, i=1,2,3, ξ_DB=0.4, P_qp=30 dBm, and the number of the sub-paths in each cluster is 8. Non isotropic omnidirectional antenna is adopted by both the transmitting antenna and the receiving antenna.

In Step S2, a V2V 2D time-domain non-stationary communication channel environment is generated by using a MATLAB, and the environment specifically includes the number of scatterers and positions of the scatterers, a random phase of a NLoS path, an angle spread of the NLoS path, a sine function lookup table, and an arctangent function lookup table.

Specifically, in this embodiment, the V2V communication channel model is adopted in the model, and the schematic diagram of the specific communication channel model is illustrated in FIG. 2.

A uniform linear array is adopted at the antenna terminal, which can be placed arbitrarily in a 2D space. For simplicity, only one cluster in each of LoS path, SB path, and DB path is shown in the drawings. Three classes of the scatterers are located on a circle with the transmitter as the center, a circle with the receiver as the center, and an ellipse representing the street environment, respectively, and only one scatterer of all three classes of the scatterers is shown in the drawings. {right arrow over (v)}_S1(2) denotes a motion velocity of scatterer in the 1^st(2^nd)class.

More specifically, in this embodiment, generating a V2V 2D time-domain non-stationary communication channel environment specifically includes the following steps.

In Step S201, the numbers of the scatterers and positions of the scatterers are generated.

An average survival probability of effective scatterers within the time period Δt is expressed as

$P_{\mathrm{remain}}=\exp \left\{-{\gamma }_R\left[P_m\left({\overrightarrow{v}}_{S_1}+{\overrightarrow{v}}_{S_2}\right)\Delta t+{\overrightarrow{v}}_{\mathrm{Rx}}-{\overrightarrow{v}}_{\mathrm{Tx}}\Delta t\right]\right\},$

where γ_R denotes a generation rate of the scatterers, P_m denotes a motion percentage, {right arrow over (v)}_S1 and {right arrow over (v)}_S2 denote an average velocity of the scatterers S₁ and an average velocity of S₂, respectively. {right arrow over (v)}_Tx and {right arrow over (v)}_Rx denote a velocity of the transmitter and a velocity of the receiver, respectively. A probability function N_new(Δt) of the number of newly generated scatterers is expressed as

$P\left(N_{\mathrm{new}}\left(\Delta t\right)=k\right)=\frac{{\lambda }^k}{k!}{e}^{-\lambda },$ $k=0,1,\dots ,$ $\mathrm{and}$ $\lambda =\frac{{\gamma }_G}{{\gamma }_R}\left(1-P_{\mathrm{remain}}\right)$

where λ denotes a parameter for a Poisson distribution probability distribution function; and γ_G denotes a disappearance rate of the scatterers. By combining the above two formulas, the average number of scatterers can be obtained:

$E\left\{N\left(t\right)\right\}=N\left(t-\Delta t\right)P_{\mathrm{remain}}+E\left\{N_{\mathrm{new}}\left(\Delta t\right)\right\}=\frac{{\gamma }_G}{{\gamma }_R}.$

Eventually, the numbers of the scatterers within each time period are obtained by calculation, which are generated through a MATLAB, and are sequentially stored in a file with a coe suffix. Position coordinates of the scatterers are uniformly and randomly generated in a rectangular area around the receiver and the transmitter, which are generated through the MATLAB, and are sequentially stored in the file with a coe suffix.

In Step S202, the random phase of a NLoS path is generated.

The random phase of a NLoS path follows a uniform distribution in [−π,π), which are generated through the MATLAB, and are sequentially stored in the file with a coe suffix.

In Step S203, the angle spread of the NLoS path is generated.

Calculation formula of the angle spread of the NLoS path is expressed as

${\phi }_{AoD}^m\left(t\right)=AS\left({\theta }_{AoD}\right)Y_{AoD}^m,$ ${\phi }_{AoA}^m\left(t\right)=AS\left({\theta }_{AoA}\right)Y_{AoA}^m,$

where AS(θ_AoD)and AS(θ_AoA)denote an angle spread coefficient for the AoD and an angle spread coefficient for the AoA, respectively, and Y_AoD^m and Y_AoA^m follow a standard normal distribution N(0,1), respectively, which are generated through the MATLAB, and are sequentially stored in the file with a coe suffix.

In Step S204, the sine function lookup table is generated.

65536 points are uniformly sampled within one cycle to be symmetric about a y-axis, and function values are amplified to 4096 times, which are generated through the MATLAB, and are sequentially stored in the file with a coe suffix.

In Step S205, the arctangent function lookup table is generated.

524288 points are uniformly sampled around an origin point with an interval 1/65536 between each sampling point to be symmetric about the y-axis, and function values are amplified to 2¹⁵/π times, which are generated through the MATLAB, and are sequentially stored in the file with a coe suffix.

In Step S3, parameters generated in Step S2 are imported into a hardware simulation platform to calculate the communication channel parameters of the clusters, such as an angle distribution and an amplitude distribution, a Verilog code is written for running, and a channel impulse response of the time-domain non-stationary V2V MIMO communication channel is eventually obtained by calculation.

In this embodiment, Step S3 specifically includes the following steps.

In Step S301, the AoD of all paths and the AoA of all paths are generated.

The calculation formulas of the AoD of all paths and the AoA of all paths are expressed as

${\theta }_{\mathrm{AoD}}^{\mathrm{LoS}}\left(t\right)=\arctan \frac{{\mathrm{Rx}}^y\left(t\right)-{\mathrm{Tx}}^y\left(t\right)}{{\mathrm{Rx}}^x\left(t\right)-{\mathrm{Tx}}^x\left(t\right)},$ ${\theta }_{\mathrm{AoD}}^{\mathrm{LoS}}\left(t\right)=\arctan \frac{{\mathrm{Tx}}^y\left(t\right)-{\mathrm{Rx}}^y\left(t\right)}{{\mathrm{Tx}}^x\left(t\right)-{\mathrm{Rx}}^x\left(t\right)},$ ${\theta }_{\mathrm{AoD}}^{i,n}\left(t\right)=\arctan \frac{S_{i,n}^y\left(t\right)-{\mathrm{Tx}}^y\left(t\right)}{S_{i,n}^x\left(t\right)-{\mathrm{Tx}}^x\left(t\right)},$ ${\theta }_{\mathrm{AoA}}^{i,n}\left(t\right)=\arctan \frac{S_{i,n}^y\left(t\right)-{\mathrm{Rx}}^y\left(t\right)}{S_{i,n}^x\left(t\right)-{\mathrm{Rx}}^x\left(t\right)},$

where S_i,n^x(t), Tx^x(t) and Rx^x(t) denote a horizontal coordinate of the n-th scatterer in the i-th class(i=1,2,3), a horizontal coordinate of the transmitter, and a horizontal coordinate of the receiver, respectively, and S_i,n^y(t), Tx^y(t) and Rx^y(t) denote a vertical coordinate of the n-th scatterer in the i-th class, a vertical coordinate of the transmitter, and a vertical coordinate of the receiver, respectively.

In Step S302, a sub-path angle of the NLoS path is generated.

The calculation formulas of a sub-path angle of the NLoS path are expressed as

${\theta }_{\mathrm{AoD}}^{i,n,m}\left(t\right)={\phi }_{\mathrm{AoD}}^m\left(t\right)+{\theta }_{\mathrm{AoD}}^{i,n},$ ${\theta }_{\mathrm{AoD}}^{i,n,m}\left(t\right)={\phi }_{\mathrm{AoA}}^m\left(t\right)+{\theta }_{\mathrm{AoA}}^{i,n},$

where φ_AoD^m(t) and φ_AoA^m(t) denote an angle spread value for the AoD of the m-th sub-path and an angle spread value for the AoA of the m-th sub-path, respectively, and θ_AoD^i,n and θ_AoA^i,n denote an average value for the AoDs of clusters passing through the n-th scatterer in the i-th class and an average value for the AoAs of the clusters passing through the n-th scatterer in the i-th class, respectively.

In Step S303, a time delay value is generated.

The calculation formula of a time delay value is expressed as

${\tau }^{\mathrm{LoS}}\left(t\right)=\frac{\sqrt{{\left[{\mathrm{Rx}}^x\left(t\right)-{\mathrm{Tx}}^x\left(t\right)\right]}^2+{\left[{\mathrm{Rx}}^y\left(t\right)-{\mathrm{Tx}}^y\left(t\right)\right]}^2}}{c}$ ${\tau }_n^{{\mathrm{SB}}_i}\left(t\right)=\frac{\sqrt{\begin{array}{c}{\left[S_{i,n}^x\left(t\right)-{\mathrm{Tx}}^x\left(t\right)\right]}^2+{\left[S_{\mathrm{in}}^y\left(t\right)-{\mathrm{Tx}}^y\left(t\right)\right]}^2+\\ {\left[S_{i,n}^x\left(t\right)-{\mathrm{Rx}}^x\left(t\right)\right]}^2+{\left[S_{i,n}^y\left(t\right)-{\mathrm{Rx}}^y\left(t\right)\right]}^2\end{array}}}{c}$ ${\tau }_{n_{1,}{n}_2}^{\mathrm{DB}}\left(t\right)=\frac{\sqrt{\begin{array}{c}{\left[S_{1,n}^x\left(t\right)-{\mathrm{Tx}}^x\left(t\right)\right]}^2+{\left[S_{1,n}^y\left(t\right)-{\mathrm{Tx}}^y\left(t\right)\right]}^2+\\ {\left[S_{1,n}^x\left(t\right)-S_{2,n}^x\left(t\right)\right]}^2+[S_{1,n}^y\left(t\right)-\\ {S_{2,n}^y\left(t\right)]}^2+[S_{2,n}^x\left(t\right)-\\ {{\mathrm{Rx}}^x\left(t\right)]}^2+{\left[S_{2,n}^y\left(t\right)-{\mathrm{Rx}}^y\left(t\right)\right]}^2\end{array}}}{c}$

where, τ^LoS(t) denotes a time delay value for a LoS path, τ_n^SBi(t) denotes a time delay value for a SB path cluster passing through the n-th scatterer in the i-th class, τ_n1,n2^DB(t) denotes a time delay value for a DB path cluster passing through the n₁-th scatterer in the first class and the n₂-th scatterer in the second class, and c denotes a velocity of light.

In Step S304, Doppler phase values are generated.

The calculation formulas of the Doppler phase values are expressed as

$f^{LoS}\left(t\right)=f_c\left(\frac{{\overrightarrow{v}}_{Tx}^T{\hat{r}}_{Tx}^{LoS}\left(t\right)}{c}+\frac{{\overrightarrow{v}}_{Rx}^T{\hat{r}}_{Rx}^{LoS}\left(t\right)}{c}\right),$

where f^LoS(t) denotes a Doppler frequency of the LoS path, f_c denotes a carrier frequency; and {circumflex over (r)}_Tx^LoS(t)=(cosθ_AoD^LoS(t),sinθ_AoD^LoS(t))

${\hat{r}}_{Rx}^{LoS}\left(t\right)=\left(\cos {\theta }_{AoA}^{Los}\left(t\right),\sin {\theta }_{AoA}^{Los}\left(t\right)\right)$ ${\varphi }^{LoS}\left(t\right)={\int }_{t_0}^{t}2\pi f^{LoS}\left(t^{\prime }\right){\mathrm{dt}}^{\prime },$

where {circumflex over (r)}_Tx^LoS(t) and {circumflex over (r)}_Rx^LoS(t) denote a unit vector for the AoD of the LoS path and a unit vector for the AoA of the LoS path, respectively, and ϕ^LoS(t) denotes a Doppler phase of the LoS path; and

$f_{n,m}^{S{B}_i}\left(t\right)=f_c\left(\frac{{\overrightarrow{v}}_{n_i,\mathrm{Tx}}^T{\hat{r}}_{\mathrm{Tx},n,m}^{\mathrm{SB}}\left(t\right)}{c}+\frac{{\overrightarrow{v}}_{n_i\mathrm{Rx}}^T{\hat{R}}_{x,n,m}^{\mathrm{SB}}\left(t\right)}{c}\right),$

f_n,m^SBi(t) denotes a Doppler frequency of the m-th sub-path of the SB path passing through the n-th scatterer in the i-th class, and {right arrow over (v)}_n,i,Tx ={right arrow over (v)}_Tx−{right arrow over (v)}_n,i

${\overrightarrow{v}}_{n_i,Rx}={\overrightarrow{v}}_{Rx}-{\overrightarrow{v}}_{n_i}$ ${\hat{r}}_{\mathrm{Tx},n,m}^{S{B}_i}\left(t\right)=\left(\cos {\theta }_{AoD}^{i,n,m}\left(t\right),\sin {\theta }_{AoD}^{i,n,m}\left(t\right)\right)$ ${\hat{r}}_{\mathrm{Rx},n,m}^{S{B}_i}\left(t\right)=\left(\cos {\theta }_{\mathrm{AoA}}^{i,n,m}\left(t\right),\sin {\theta }_{\mathrm{AoA}}^{i,n,m}\left(t\right)\right)$ ${\varphi }_{n,m}^{S{B}_i}\left(t\right)={\int }_{t_0}^{t}2\pi f_{n,m}^{S{B}_i}\left(t^{\prime }\right){\mathrm{dt}}^{\prime },$

where {right arrow over (v)}_ni,Tx and {right arrow over (v)}_ni,Rx denote a relative velocity of the transmitter with a n-th scatterer and a relative velocity of the receiver with the n-th scatterer, respectively; {circumflex over (r)}_Tx,n,m^SBi(t) and {circumflex over (r)}_Rx,n,m^SBi(t) denote a unit vector for an AoD of the m-th sub-path of the SB path passing through the n-th scatterer in the i-th class, and a unit vector for an AoA of the m-th sub-path of the SB path passing through the n-th scatterer in the i-th class, respectively, and ϕ_n,m^SBi(t) denotes a Doppler phase of the m-th sub-path of the SB path passing through the n-th scatterer in the i-th class;

$f_{n_1,n_2,m}^{\mathrm{DB}}\left(t\right)=f_c\left(\frac{{\overrightarrow{v}}_{n_i,\mathrm{Tx}}^T{\hat{r}}_{\mathrm{Tx},n_1,m}^{\mathrm{DB}}\left(t\right)}{c}+\frac{{\overrightarrow{v}}_{n_i\mathrm{Rx}}^T{\hat{r}}_{x,n_2,m}^{\mathrm{DB}}\left(t\right)}{c}\right),$

where f_n1,n2,m^DB(t) denotes a Doppler frequency of the m-th sub-path of the DB path cluster passing through the n₁-th scatterer in the first class and the n₂-th scatterer in the second class; and

${\overrightarrow{v}}_{n_1,\mathrm{Tx}}={\overrightarrow{v}}_{Tx}-{\overrightarrow{v}}_{n_1}$ ${\overrightarrow{v}}_{n_2,Rx}={\overrightarrow{v}}_{Rx}-{\overrightarrow{v}}_{n_2}$ ${\hat{r}}_{\mathrm{Tx},n_1,m}^{DB}\left(t\right)=\left(\cos {\theta }_{AoD}^{1,n_1,m}\left(t\right),\sin {\theta }_{AoD}^{1,n_1,m}\left(t\right)\right)$ ${\hat{r}}_{\mathrm{Rx},n_2,m}^{DB}\left(t\right)=\left(\cos {\theta }_{AoA}^{2,n_2,m}\left(t\right),\sin {\theta }_{AoA}^{2,n_2,m}\left(t\right)\right)$ ${\varphi }_{n_1,n_2,m}^{DB}\left(t\right)={\int }_{t_0}^{t}2\pi f_{n_{1,}{n}_2,m}^{\mathrm{DB}}\left(t^{\prime }\right){\mathrm{dt}}^{\prime },$

where {circumflex over (r)}_Tx,n1,m^DB(t), and {circumflex over (r)}_Rx,n2,m^DB(t) denote a unit vector for an AoD of the m-th sub-path of a cluster of the DB path passing through the n₁-th scatterer in the first class and the n₂-th scatterer in the second class, and a unit vector for an AoA of the m-th sub-path of the cluster of the DB path passing through the n₁-th scatterer in the first class and the n₂-th scatterer in the second class, respectively, and ϕ_n1,n2,m^DB(t)denotes a Doppler phase of the m-th sub-path of the cluster of the DB path cluster passing through the n₁-th scatterer in the first class and the n₂-th scatterer in the second class.

In Step S305, an antenna phase value is generated.

The calculation formula of the antenna phase value is expressed as follows.

A relative coordinate {right arrow over (d)}_Tx of a transmitter antenna is:

${\overrightarrow{d}}_{Tx}={\left[d_{\mathrm{Tx}}^x,d_{Tx}^y\right]}^T,$

where d_Tx^x denotes a horizontal ordinate of {right arrow over (d)}_Tx and d_Tx^y denotes a vertical ordinate of {right arrow over (d)}_Tx

A phase difference ψ_Tx(t) of the transmitter antenna is:

${\psi }_{Tx}\left(t\right)=2\pi f_c\frac{{\hat{r}}_{\mathrm{Tx}}^T\left(t\right)·{\overrightarrow{d}}_{\mathrm{Tx}}}{c},$

where {circumflex over (r)}_Tx(t) denotes a unit vector of an AoD at the time instant t.

A relative coordinate {right arrow over (d)}_Rx of the receiver antenna i:

${\overrightarrow{d}}_{Rx}={\left[d_{\mathrm{Rx}}^x,d_{Rx}^y\right]}^T,$

where dix denotes a horizontal ordinate of {right arrow over (d)}_Rx, d_Rx^y denotes a vertical ordinate of {right arrow over (d)}_Rx. A phase difference ψ_Tx(t) of the receiver antenna is:

${\psi }_{Rx}\left(t\right)=2\pi f_c\frac{{\hat{r}}_{\mathrm{Rx}}^T\left(t\right)·{\overrightarrow{d}}_{Rx}}{c},$

where {circumflex over (r)}_Rx(t) denotes a unit vector for an AoA at the time instant t.

The calculation formula of the total antenna phase value ψ(t) is expressed as:

$\psi \left(t\right)={\psi }_{\mathrm{Tx}}\left(t\right)+{\psi }_{\mathrm{Rx}}\left(t\right).$

In Step S306, amplitude values are generated.

The calculation formulas of the amplitude values are expressed as:

$H^{LoS}\left(t\right)=\sqrt{\frac{K_{qp}{P}_{qp}}{K_{qp}+1}}$ $H^{S{B}_i}\left(t\right)=\sqrt{\frac{{\xi }_{S{B}_i}{P}_{qp}}{\left(K_{qp}+1\right)N_i\left(t\right)M}}$ $H_{qp}^{DB}\left(t\right)=\sqrt{\frac{{\xi }_{DB}{P}_{qp}}{\left(K_{qp}+1\right)N_1\left(t\right)N_2\left(t\right)M}},$

where, H_qp^LoS(t) denotes channel impulse response amplitude value for a LoS path between the q-th receiving antenna and the p-th transmitting antenna, H_qp^SBi(t) denotes a channel impulse response amplitude value for a sub-path of a SB path passing through the scatterer in the i-th class located between the q-th receiving antenna and the p-th transmitting antenna, H_qp^DB(t) denotes a channel impulse response amplitude value for the sub-path of the DB path between the q-th receiving antenna and the p-th transmitting antenna, K_qp denotes a Rice factor of a p-q link, P_qp denotes a total power of the p-q link, ξ_SBi and ξ_SBi denote the proportion of the total scattering power of the SB ray and the DB ray on the NLoS path, respectively, N_i(t) denotes the number of the scatterer in the i-th class at the time instant t, and M denotes the number of the sub-paths in each cluster of the NLoS path.

K_qp, ξ_SBi and ξ_DB in the model are related to the macro cells, the micro cells, and the micro micro cells, as well as a SB path and a DB path. The model proposed by the present disclosure can adapt to various V2V propagation environments by adjusting the model parameters ξ_SBi, ξ_DB and the Rice factor K_qp. In the macro cell scenarios, due to a relative large distance between the transmitter and the receiver, the DB rays carry more energy than that of the SB rays(the relative large distance between the transmitter and the receiver leads to a greater independence of an AoD and an AoA, that is, ξ_DB>max{ξ_SB1,ξ_SB2}>>ξ_SB3, and the received signal power mainly comes from the SB rays and the DB rays of the dual loop model. Therefore, the Rice factor K_qp and the energy parameter ξ_SB3 are extremely little, even close to zero. This means that in the macro cell scenario, it can be characterized by using a dual loop model that ignores the LoS radial component. Compared with the macro cell scenario, in the micro cell scenario and the micro-micro cell scenario, VTD significantly affects the channel characteristics. In consideration of the impacts of VTD on channel statistics, the present disclosure distinguishes the mobile cars and the stationary roadside environments(such as buildings, trees, parked cars, etc.)around the transmitter and the receiver. Therefore, the present disclosure simulates a mobile car by using a dual loop model and describes a stationary roadside environment by using an elliptical model. For low VTD, since the LoS path has a high power, the value for K_qp is large. In addition, the received scattering power mainly comes from the signal reflected by the stationary roadside environment described by the scatterers freely located on the ellipse. Since the mobile car represented by the scatterers located on the dual loop is sparse, it is more likely to be a SB rather than a DB, indicating a validity of ξ_SB3>max{ξ_SB1,ξ_SB2}>ξ_DB is. Under a condition of a high VTD, the value for K_qp is smaller than that in a low VTD scenario, and due to the relative large number of the mobile cars, the DB rays of the dual loop model carry more energy than those of the SB rays of the dual loop model and the elliptical model, that is, ξ_DB>max {ξ_SB1,ξ_SB2,ξ_SB3}. Thus, the micro cell scenario and the micro-micro cell scenario in consideration of VTD can be well characterized by utilizing the combined dual loop model and the elliptical model with a LoS radial component.

In Step S307, channel impulse response is generated.

The calculation formula of the channel impulse response is expressed as

$h_{qp}^{LoS}\left(t,\tau \right)=H^{LoS}\left(t\right)e^{-j2\pi f_c{\tau }^{LoS}\left(t\right)}{e}^{j{\varphi }^{{\mathrm{LoS}}_{\left(t\right)}}}{e}^{j{\psi }^{LoS}\left(t\right)}\delta \left(\tau -{\tau }^{LoS}\left(t\right)\right)$ $h_{qp}^{SB}\left(t,\tau \right)=\sum _{i=1}^3\sum _{n_i=1}^{N_i\left(t\right)}{H}^{S{B}_i}\left(t\right)e^{-j2\pi f_c{\tau }_n^{S{B}_i}\left(t\right)}{e}^{j{\phi }_{n,m}^{S{B}_i}\left(t\right)}{e}^{j{\varphi }_{n,m}^{S{B}_i}\left(t\right)}{e}^{j{\psi }_{n,m}^{S{B}_i}\left(t\right)}\delta \left(\tau -{\tau }_n^{S{B}_i}\left(t\right)\right)$ $h_{qp}^{DB}\left(t,\tau \right)=\sum _{n_1,n_2=1}^{N_1\left(t\right),N_2\left(t\right)}{H}_{n_1,n_2,m}^{\mathrm{DB}}\left(t\right)e^{-j2\pi f_c{\tau }_{n_1,n_2}^{DB}\left(t\right)}{e}^{j{\phi }_{n_1,n_2,m}^{DB}\left(t\right)}{e}^{j{\varphi }_{n_1,n_2,m}^{DB}\left(t\right)}{e}^{j{\psi }_{n_1,n_2,m}^{DB}\left(t\right)}\delta \left(\tau -{\tau }_{n_1,n_2}^{DB}\left(t\right)\right)$ $h_{qp}\left(t,\tau \right)=h_{qp}^{LoS}\left(t,\tau \right)+h_{qp}^{SB}\left(t,\tau \right)+h_{qp}^{DB}\left(t,\tau \right),$

where h_qp^LoS(t,τ) denotes a channel impulse response for the LoS path between the q-th receiving antenna and the p-th transmitting antenna, h_qp^SB(t,τ) denotes the channel impulse response for the SB path between the q-th receiving antenna and the p-th transmitting antenna, h_qp^DB(t,τ) is a channel impulse response for the DB path between the q-th receiving antenna and the p-th transmitting antenna, f_c denotes a carrier center frequency, τ denotes a time delay, ϕ denotes a Doppler phase, ψ denotes an antenna phase difference, φ_n,m^SBi(t) and φ_n1,n2,m^DB(t) denote a random phase of the m-th SB sub-path passing through the n-th scatterer in the i-th class and a random phase of a m-th SB sub-path passing through the n₁-th scatterer in the first class and the n₂-th scatterer in the second class, respectively; the channel impulse response is imported from the file with a coe suffix and a calculated channel impulse response is exported to a txt file.

In Step S4, a comparison is performed with a statistical characteristic of a theoretical communication channel model, and an appropriate hardware diagram of a communication channel emulator is designed.

The specific hardware block diagram is illustrated in FIGS. 3 and 4. In this embodiment, Step 4 specifically includes the following steps.

In Step S401, a calculation formula of a time autocorrelation function TACF curve is expressed as:

$r_{\mathrm{qp},q^{\prime }{p}^{\prime }}=E\left\{{h_{qp}\left(t,\tau \right)\left[h_{q^{\prime }{p}^{\prime }}\left(t,\tau \right)\right]}^{*}\right\},$

where h_qp(t,τ)denotes a channel impulse response between the q-th receiving antenna and the p-th transmitting antenna in a case where a time period is t and a time delay is τ, h_q′p′(t,τ) denotes a channel impulse response between the q′-th receiving antenna and the p′-th transmitting antenna in a case where a time period is t and a time delay is τ; (·)* denotes a conjugate complex of(·).

In Step S402, a calculation formula of a SCCF curve is expressed as

$r_{qp}\left(\Delta t\right)=E\left\{{h_{qp}\left(t,\tau \right)\left[h_{qp}\left(t+\Delta t,\tau \right)\right]}^{*}\right\}.$

In Step S403, a calculation formula of a delay PSD curve is expressed as

${\rho \left(\tau ,{\theta }_{AoA}^{i,n}\right)={h_{qp}\left(t,\tau \right)}^2❘}_{{\theta }_{\mathrm{AoA}}^{i,n}}.$

In order to verify the correctness of the method provided in this embodiment, experiments are conducted, and specifically lie as follows.

The TACF between the transmitting antenna and the receiving antenna at different time intervals are verified, and the results are as illustrated in FIG. 5. As the time interval increases, the TACF of the NLoS path is a constant, while the TACF of the SB path and DB path decrease with the increase of the time intervals. And the reason of that is that the time changes cause the changes on the angular distributions of sub-path and the Doppler phase.

The SCCF between the receiving antennas at different antenna intervals are verified, and the results are as illustrated in FIG. 6. As the antenna interval increases, the SCCF of the NLoS path is a constant, while the SCCF of the SB path and DB path decrease with the increase of the antenna intervals. And the reason of that is that the changes in antenna intervals cause changes on the phase differences of the antenna array in the sub-path.

The delay PSD of channel impulse response at different antenna intervals are verified, and the results are as illustrated in FIG. 7 and FIG. 8. Observing the drawings, the hardware simulation results match well with the MATLAB simulation results. The coherent lines in FIG. 7 and FIG. 8 represent the delay PSD of the channel impulse response of cluster passing through the scatterers in the first class in the LoS path and SB path. This is because the scatterers in the first class is relative closer to the transmitter and relative farther from the receiver. Therefore, even if one part of the scatterers in the first class is generated or disappears, the AoA is basically the same as the angle from the transmitter to the receiver. The other scatter points in FIG. 7 and FIG. 8 indicate that as the time delay increases, the clusters of other NLoS paths are generating or disappearing. It can be seen from the delay PSD drawings, this embodiment has a characteristics of birth-death process of clusters in time-domain, which is an important characteristic of time-domain non-stationary communication channels.

In summary, provided in the present disclosure is a method for designing a time-domain non-stationary V2V MIMO communication channel emulator, which can describe the fading characteristics of macro cell, micro cell, and micro-micro cell, as well as the birth and death process of the clusters. The statistical characteristics of the simulations have important reference value for the designs of time-domain non-stationary communication channel emulators.

The unspecified parts in the present disclosure are all the common sense for a person skilled in the art. The preferred specific embodiments of the present disclosure are described in details in above. It should be understood that various amendments and changes can be made by an ordinary person skilled in the art according to the concept of the present disclosure with no creative efforts. Therefore, all technical solutions that can be obtained by a person skilled in the art on a basis of the prior art according to the concept of the present disclosure through the logical analysis, reasoning, or limited experiments should be within the protection scope determined by the claims.

Claims

1. A method for designing a time-domain non-stationary V2V MIMO communication channel emulator, wherein the method comprises following steps: Step S1, determining basic parameters for the V2V MIMO communication channel;Step S2, generating, by using a MATLAB, a V2V 2D time-domain non-stationary communication channel environment, specifically including: a number of scatterers and positions of the scatterers, a random phase of a NLoS path, an angle spread of the NLoS path, a sine function lookup table, and an arctangent function lookup table;Step S3, importing parameters generated in Step S2 into a hardware simulation platform to calculate communication channel parameters for clusters, wherein the communication channel parameters include an angle distribution and an amplitude distribution, writing a Verilog code for running, and eventually calculating to obtain a channel impulse response of the time-domain non-stationary V2V MIMO communication channel; andStep S4, comparing with a statistical characteristic of a theoretical communication channel model, and designing an appropriate hardware diagram of a communication channel emulator.

2. The method for designing the time-domain non-stationary V2V MIMO communication channel emulator according to claim 1, wherein in Step S1, in a geometry-based stochastic model for the time-domain non-stationary V2V MIMO communication channel, the basic parameters for the V2V MIMO communication channel include a number of a simulation time point, a simulation time interval, a position of a transmitter, a position of a receiver, a velocity of the transmitter, a velocity of the receiver, a Rice factor, an angle spread coefficient, a total link power, a LoS path, a power ratio of a SB path, a power ratio of a DB path, a number of initial scatterers, a velocity of the initial scatterers, a number of sub-paths in each cluster, a generation rate for scatterers, a disappearance rate for the scatterers, a motion ratio of the scatterers, a relative coordinate of a receiver antenna and a relative coordinate of a transmitter antenna.

3. The method for designing the time-domain non-stationary V2V MIMO communication channel emulator according to claim 1, wherein Step S2 specifically includes: Step S201, generating the numbers of the scatterers and the positions of the scatterers, and expressing an average survival probability Premain of effective scatterers within a time period Δt as

4. The method for designing the time-domain non-stationary V2V MIMO communication channel emulator according to claim 1, wherein in Step S3, the angle distributions of the clusters are expressed as θAoDLoS(t), θAoALoS(t), θAoDi,n(t)and θAoAi,n(t), the amplitude distribution is expressed as HqpLoS(t), HqpSBi(t) and HqpDB(t); wherein an angle of the clusters is determined by the coordinate of the transmitter, the coordinate of the receiver, and the coordinates of the scatterers; the amplitude is determined by the Rice factor, the total power, a proportion of a SB ray and a DB ray to a total scattering power on the NLoS paths, a number of an i-th class scatterers at a time instant t, and the number of sub-paths in each cluster of the NLoS path, Step S3 specifically includes: Step S301, generating the AoD of all paths and the AoA of all paths, and expressing calculation formulas of the AoD and the AoA as

5. The method for designing the time-domain non-stationary V2V MIMO communication channel emulator according to claim 1, wherein in Step S4, formulas for the statistical characteristic specifically include: Step S401, expressing a calculation formula of a time autocorrelation function TACF curve as: