The Research On The Non-stationary Properties Of Channel Modeling And Simulation Analysis

In wireless communication systems, the channel is the physical medium that connects with the transmitter and receiver. And its characteristics directly determine the performance of the wireless communication system, such as transmission capacity, bits error rate, and throughput. While due to the interaction of electromagnetic wave reflection, diffraction and scattering mechanism, the simulation of an accurate and effective channel models has been proven to be an important step on testing of communication system. But the majority of existing channel models in the researches rely on the assumption of WSSUS (Wide-Sense Stationary Uncorrelated Scattering). That's to say, the physical properties of transmission channel are relatively stationary within a very short period.With the rapid development of next generation wireless networks, especially the wide application of 5G technology, Plenty of empirical and analytical reports reveal that the channel model based on WSSUS assumption cannot completely describe the actual environments and the non-stationary properties of channel are indispensable for physical layer design and performance validation. Those findings have attracted more and more attention between the industry and academia. Therefore in this paper, the research is carried out deeply aimed at the problems of the non-stationary properties on channel modeling. In particular, it takes the newly 5G technology and scenarios (such as V2V, HST and Massive MIMO ) into consideration which makes the research results is very valuable for the simulation, evaluation, performance optimization and network deployment of future wireless mobile communication systems and provides strong support for the construction of country communication industry. In conclusion, the main works are as follows:1. V2V (Vehicle-to-vehicle) three-dimensional non-stationary channel modeling. In this part of study, we provide a novel three-dimensional (3D) non-stationary channel model, in which the scatterers are assumed to be distributed around the transmitter and receiver on a two-sphere model. By employing the von Mises Fisher distribution, the mean values of scatter angle are tracked by time-variant Brownian Markov random motion paths, which keeps the non-stationarity of the proposed channel model. Furthermore, the auto-correlation function (ACF)of the channel gain and local power spectrum density (PSD) of the Doppler frequencies are calculated by non-stationary signal processing tools, such as STFT and FFT. At last, the simulation analysis shows that the time-variant (TV)scatterer distribution results in a non-stationary non-isotropic channel model.2. HST (High Speed Train) non-stationary channel modeling. In this part,we propose a kind of geometry-based stochastic channel model by exploiting sum-of-chirps (SoC) process method. Especially, the proposed channel model accounts for the time-variant characteristics of channel due to the changes of train's velocities and trajectories. Furthermore, the TV auto-correlation function(ACF) and Doppler power spectrum density (DPSD) of the proposed channel model are derived by employing Wigner signal analysis technology. In addition to reproduce the non-stationary characteristic perfectly, we develop a low complexity parameter computation approach with linear variation.3. Massie MIMO non-stationary channel modeling. In this part of research,a new 3D massive MIMO channel model is developed by taking into account the antenna effects (such as mutual coupling and antenna polarization) and propagation effects such as scattering angles, the birth-death process of clusters and channel depolarization. In addition, the channel correlation of different polarization combinations is evaluated and the performance of massive MIMO systems deploying different antenna arrays is investigated. Theoretical analysis and numerical simulation show that the spatial-temporal properties of the proposed channel model fit perfectly to real-world scenarios.4. Channel model simulator and Emulator. In this part, we propose a generalized and accurate algorithm to generate Nakagami channels with arbitrary temporal-autocorrelation and spatial-crosscorrelation. Specifically the temporal-autocorrelation of Nakagami channels is derived using rank matching technology,then spatial-crosscorrelation is introduced by Sim's algorithm. In addition, this paper also designs and develops the framework of the generic radio channel models emulation based on the standard simulation procedure of Geometry-Based Stochastic Model (GBSM), such as 3GPP SCM (3rd Generation Partnership Project Spatial Channel Model), SCME (Spatial Channel Model Extend), WINNER (Wireless World Initiative New Radio), WINNER-? and ITU-R (International Telecommunication Union-Radiocommunication Sector)M.2135,TR.36.873.In summary, this paper faces the future 5G communication system, focus on the shortcomings and deficiencies of existing WSSUS channel model, proposed a full set of non-stationary properties channel models. Finally, on the basis of theoretical modeling, the implementation and verification of the related hardware emulation algorithms are also discussed.