Research On Modeling And Simulating Space-time-frequency Correlated MIMO Channel Under Non-isotropic Scattering

MIMO technology can be used to improve system capacity, transmission rate and spectralefficiency and becomes one of the key solutions for the4thgeneration wireless communication system.It is well known that the performance of MIMO system is greatly influenced by the wirelesspropagation environment. The correlation between channel fadings cannot be ignored for the compactantenna array and non-ideal propagation. Therefore, it is very important for parameters design,algorithms optimization and performance analysis of MIMO system to study the modeling and thesimulating MIMO channels with space-time-frequency correlation. This dissertation focuses oncorrelated MIMO channel model in non-isotropic scattering environment, space-time-frequencycorrelation between MIMO signals and MIMO system performance analysis for composite fading.Firstly, the wireless propagation properties and the researches on channel model and simulationare reviewed. The stability and statistically independence of the traditional SoS models are not good,so an improved model based on optimal angle of arrival (AoA) is proposed, whose statisticalperformance is better than other models. Based on the method of equal probability areas of AoA, anextended SoS model for non-isotropic scattering conditions is presented. It is then applied onmodeling the Nakagami fading channel which is more flexible than Rayleigh fading. Thus it lays thefoundation of Nakagami-MIMO channel model.Secondly, this paper analyzes the space-time-frequency correlations of MIMO system underRayleigh fading and derives the expression of spatial-temporal joint correlation with mobile antenna.The evaluation of spatial correlation is very important but also complicated for needing numericalcomputation, so a simple spatial correlation calculation method based on pulse sampling principle inangle domain is presented. The maximum error of the new method is less than0.01while thecomputation time is far less than numerical integration method or other methods. Based on therelationships between Rayleigh and Nakagami fading, a general close-form expression of thespace-time-frequency correlation for Nakagami fading, as well as coherence time and coherencebandwidth, are derived. These results are useful and important for MIMO channel model andperformance analysis.Thirdly, based on the SoS model proposed and correlation properties derived, this paper thenpresents a new correlated MIMO channel model with Nakagam fading. It obtains the channelparameters from field data or statistical information, and then calculates the space-time-frequencycorrelation matrix by theoretical method. Finally, the Nakagami-MIMO channels are obtained by generating Nakagami stochastic processes with specific auto-correlation and cross-correlation. Inorder to realize the large-scale and real-time MIMO channel simulation, a new rejection method isalso presented for generating Nakagami sequences with arbitrary fading factor, which has higherefficiency than other methods reported.Finally, it is very difficult to analyze the system performance with the complicated distribution ofcomposite fading, so an approximate method based on moments match is proposed. In order toconfirm the validity of the new method, a novel simulation model for composite fading channel isthen presented. Moreover, it is applied on getting the ergodic capacity of correlated composite fadingMIMO channels.