| The demand for the system capacity and mobile data rate is expected to increase sharply in current and future mobile communication systems.To satisfy such demand,new frequency bands will be used,the spectrum efficiency should be increased significantly as well.Thus,accurate channel modelling for new frequency bands and mobile scenarios is crucially important.Channel models,either analytical models or physical models,are developed with the extracted parameters from the channel measurement campaign or the channel simulation.With the expansion of the wireless scenarios and the spectrum resources,the manual and financial cost of the onsite channel measurement will increase dramatically or even be impossible.On the other hand,under the improvement of the simulation algorithm and the increment of computing power,it is more convenient to use the electromagnetic simulation for channel modelling.There are two typical electromagnetic methods for the channel simulation: the ray-tracing method and the Finite-Difference Time-Domain(FDTD)method.The ray-tracing method is an approximate solution based on the hypothesis of high-frequency electromagnetic waves.The path can be intuitively displayed and the simulation is fast.However,due to the hypothesis,the application of the raytracing method is limited.The FDTD method is a full-wave method based on Maxwell’s curl equations.It can provide highly accurate results.However,for the large-scale wireless channel,too many computational resources are required.In this dissertation,the FDTD method is used in large-scale wireless channels.To achieve high accuracy and simulation speed,many improvements are developed.For the first time,the extraction method of all parameters needed in the wireless system is proposed with the FDTD method; For the first time,the Monte Carlo method is used in the FDTD method to solve the problem of channel statistical characteristics; For the first time,the frequency transformation method is proposed to solve the problem of large-scale channels in higher frequency; For the first time,the complex envelope leapfrog alternating-direction-implicit FDTD method is developed to solves the problem of broadband digital modulated signal transmission simulation.These developed methods have been successfully employed in several wireless scenarios requiring accurate channel data,and good agreement with the experimental results is achieved.The detailed contents and innovations of this dissertation are as follows:1)According to the channel models used in the wireless system,a novel method to extract the parameters of the practical application scenarios with the FDTD simulation is developed.All the important channel parameters,including the multipath,the path decay and time delay,arrival angle calculation,transceiver angle pairing,etc.,can be determined accurately.Compared with the ray-tracing method,this method can be used in a wide range of applications and has a higher calculation accuracy.With this method,the path loss and time delay of the typical full-duplex channel are extracted and applied in the self-interference cancellation circuit.Excellent agreement is achieved.2)According to the stochastic feature of the wireless channel,for the first time,the Monte Carlo method is introduced into the FDTD method.Most of the important channel parameters,such as the path loss,the propagation delay,the multipath characteristics,are obtained and the statistics is presented for a specific channel.This method is used to simulate the 3D wireless channel in vertical direction successfully.3)To reduce the requirement of simulation consumption in the millimetre-wave channels and the large-scale three-dimensional channels,two fast channel simulation methods are proposed: one is the characteristic plane method and the other is the frequency transformation method.In the characteristic plane method,several optimal planes are selected for the 2D simulation.With these 2D simulated results,the 3D channel parameters can be synthesized with a certain accuracy.In the frequency transformation method,according to the theoretical analysis of the reflection,many high-frequency channels can be transformed to lower frequency bands by adding some compensation.These methods are verified with the full 3D simulation and good agreement is obtained.4)To simulate the wireless channel with the requirement of very high accurate data analysis,such as the OTA test,for the first time,the complex envelope leapfrog alternating-direction-implicit Finite-Difference Time-Domain method is proposed.This method is unconditionally stable and has an extremely high accuracy near the centre frequency.It is suitable for the OTA symbol-level time simulation.By loading the baseband envelope on the transmitter,the whole OTA RF test link is simulated,and the performance of the EVM test is accurately obtained.This method can quantify the impact of measurements at different reflectance coefficients for nonideal chambers.The examples fully demonstrate the stability and high accuracy of the proposed algorithm.Simulation results agree well with the experiment.In summary,novel methods are developed for the wireless channel parameter extraction with the FDTD method,good accuracy and high simulation efficiency is obtained.The methods proposed in this dissertation has been successfully used in many national programs and university-enterprise cooperation programs.The work has been published in IET journals and IEEE international conferences.Some of the work has been applied for national patents. |