Research On Non-stationary Channel Modeling And Space-time-frequency Characteristics Of Vehicular Communications

At the end of 2020,the state council,together with 11 ministries and committees,jointly pointed out that building an efficient and reliable vehicular communication system based on cellular vehicle-to-everything(C-V2X)technology to promote the development of the intelligent and connected vehicle industry is a critical issue to realize China’s industrial transformation and upgrading,and improve people’s well-being.To establish an efficient and reliable vehicular communication system,it is vital to master the channel characteristics in detail,which can be obtained through channel measurement and channel modeling.According to the reliable channel characteristics,the reasonable configuration of transceivers and corresponding algorithms can be further designed to achieve the optimal system performance.Therefore,in addition to channel measurement campaigns,it is essential to study channel modeling and its space-time-frequency correlation characteristics,especially for non-stationary channel modeling in high-speed scenarios.Combined with multiple-input multiple-output(MIMO)technology,this thesis adopts the geometry-based stochastic model(GBSM)to conduct an in-depth study on the non-stationary channel modeling and simulation analysis of wideband channels in moving scenarios.The main innovation points are as follows:(1)In order to effectively simulate the real communication environment and characterize the channel non-stationarity,a non-stationary wideband MIMO channel model which combines multiple ellipses model and one-ring model with time-varying parameters is proposed in this thesis,and the mathematical expression of the channel impulse response(CIR)is given in detail.In the proposed channel model,the stationary scatterers and the moving scatterers in the propagation environment are respectively represented by the ellipse model and the one-ring model,and the receiver and the moving scatterers move in any direction with a time-varying velocity and a constant velocity,respectively.Then,the birth and death processes are introduced to consider the different numbers of ellipses at different time instants.According to the geometry relationship of the proposed channel model and the time evolution process,the updated expressions of the time-varying parameters are derived respectively,such as the number of ellipses,Doppler frequency shifts,propagation distances,angle of departures(Ao Ds),angle of arrivals(Ao As)and the delays of different paths.(2)In order to reduce the computation complexity,a corresponding stochastic simulation model is developed based on the reference model with the finite number of scatterers,and its CIR expression is given.In this stochastic simulation model,the set of initial values of the discrete AODs or AOAs can be obtained by using a reasonable parameter calculation method and then updated according to the procedures in the reference model.(3)Based on the proposed non-stationary channel reference model and stochastic simulation model,the expressions of spatial cross-correlation function(S-CCF),temporal auto-correlation function(T-ACF)and frequency correlation function(FCF)are derived,and then the corresponding numerical simulation analysis is carried out respectively.The influence of different parameters on channel characteristics are investigated,such as antenna spacings,antenna tilt angles,moving velocity and direction,time-varying angles,the parameters concerning the scattering environment and vehicle traffic density(VTD).Numerical results show that the proposed channel model can imitate the channel non-stationarity in the time domain,and the channel correlation statistical characteristics have significant differences under different VTD scenarios.Meanwhile,the high consistency of the numerical results between the reference model and the simulation model at different time instants verifies the accuracy of the theoretical derivation and the practicability of the simulation model.The proposed channel model can be applied to different communication scenarios by changing the model-related and environment-related parameters.Additionally,the numerical results also provide some reference for vehicular communication systems,for example,the antenna configuration concerning different moving directions.