Research On Ray Tracing Simulation And Modeling Of Millimeter-wave Mobile Channels

With large available spectrum resources,millimeter-wave(mm Wave)communication has become a key technology to support high-data-rate,low-latency,and ultra-reliable transmission in 5G and beyond mobile communication systems.Compared to the sub-6GHz bands,mm Wave signals are of high free space loss,serious blockage attenuation,determinism,and strong spatial consistency in mobile communication scenarios.These characteristics impose stricter requirements on channel modeling and make it more challenging.Though mm Wave propagation characteristics can be obtained by channel measurements,it requires expensive equipment and complex deployment process.More importantly,it is impossible to measure the complex mobile scenarios dynamically.Based on the physical simulation model of radio wave propagation mechanism,ray tracing(RT)can simulate the space,time,frequency and polarization information in the channel,which has been an important technology to break through the limitation of channel measurements.Yet this method has been adopted by 3rd generation partnership project(3GPP)and international telecommunication union(ITU)standard organizations.However,due to the high computational complexity and long simulation time of RT,it is still challenging to simulate channels in complex geometric scenarios.How to enhance the RT simulation efficiency for complex mobile scenario is an urgent issue.Oriented by the modeling requirements of mm Wave mobile channels,this dissertation constructs an efficient RT simulation platform based on the space partition theory and graphics processing unit(GPU)parallel computing technology.The material parameters in the propagation model are calibrated using channel measurement data and genetic algorithms.Based on the efficient and accurate RT simulator,a detailed and comprehensive three-dimensional(3D)radio wave propagation simulation is carried out for vehicle-to-infrastructure(V2I)and train-to-infrastructure(T2I)communication scenarios.The influence of significant scatterers and the evolution of key characteristics in the channel are extracted and analyzed.Finally,by leveraging the semi-deterministic modeling method,a spatially consistent mobile channel model conforming to the evolution property is modelled for V2 I scenarios.The main contributions of this dissertation are as follows.1.To enhance the ray tracing simulation efficiency,this dissertation proposes a Hashbased dynamic uniform grid space partition algorithm for mobile communication scenarios,which enhances the construction efficiency of acceleration structures and the speed of ray intersection detection,making it possible to simulate accurate radio wave propagation with the consideration of massive moving objects.Moreover,based on the real-time ray tracing technology of GPU,this dissertation designs a simulation method of radio wave propagation by the rendering pipeline of GPU.According to the benchmark results in complex scenarios,the intersection detection speed of this method is 1000 times faster than that of CPU,which further enhances the simulation efficiency of the RT model.2.To investigate the characteristics of the V2 I mm Wave mobile channel,the propagation model of RT is calibrated by the genetic algorithm with measurement data.Subsequently,according to the vehicle mobility model in 3GPP TR 37.885,the 3D RT simulations are performed in an urban road environment,where vehicles run at different speeds.Next,channel characteristics such as blockage probability,power delay profile,path loss,delay spread,and angular spread are analyzed and modeled comprehensively.The influence of scatterers on the channel and the evolution of channel parameters are revealed as well,which provides a basis for modeling a spatially consistent mobile channel model.3.To investigate the characteristics of T2 I mm Wave mobile channels,this dissertation conducts comprehensive simulation of different deployments in rural railway scenario with calibrated RT model.The space-time-frequency multi-dimensional statistical characteristics of the channel are extracted from the simulations.Then,the evolution characteristics of each channel parameter are studied,and evolution models for delay spread,Rician K-factor,and angular spread are expressed as continuous or piecewise functions.In addition,the decorrelation distance of channel parameters is studied,and the variation law and reason for the number of domain multi-paths are revealed.Such complete channel parameters can be directly used for 3GPP-like stochastic channel models.4.To study the spatial consistency modeling method of mm Wave mobile channels,this dissertation conducts intensive simulations in V2 I scenarios based on the calibrated RT model.The line-of-sight(LOS)probability model under different vehicle speeds and transceiver heights is extracted.Moreover,a simplified geometric model is used to model the communication scenario,and a deterministic reflection model is established according to the geometric features.In addition,a stochastic reflection model is used to model the reflection paths that cannot be described by geometry;a reflection path loss model related to the distance and reflection times is established;the evolution characteristics of multipath number and angle spread are analyzed.This dissertation builds a spatially consistent V2 I channel model using geometrically correlated deterministic reflection and scattering models,stochastic reflection models,and probabilistic blocking models.