The Research On The Validity Of Multiple Antennas Communication System With Spatial And Temporal Correlation

Multiple Input Multiple Output(MIMO) system can offer large gains in capacity for wireless communication system. Therefore, as a means of higher transmission data rate, it has received considerable attention. However, the application of MIMO system is confined for the spatial and temporal correlation. The thesis concerns the validity of MIMO system with spatial and temporal correlation via analyzing the MIMO channel model. The main contribution is described as follows:Firstly, on the basis of investigating the MIMO channel model, we propose an effective cluster-based geometric model for MIMO channel, and deduce the spatial and temporal correlation coefficient. Then, the correlated capacity of flat Rician fading channel is analyzed. The effect of different parameters on the ergodic and outage capacity is performed by theoretic analysis and simulations.Secondly, with effective clusters, a geometric MIMO channel model for frequency-selective Rician fading is presented. The capacity of the MIMO-OFDM system is derived from this model with double correlation scenarios. The effect of different parameters on the frequency-selective capacity is investigated. The relation between multiple clusters in flat fading and multiple taps in frequency-selective channel is analyzed, and conclude that the two is spatial and temporal duality.Thirdly, according to the flat Rician fading model we have proposed, the capacity bounds are validated, and the bounds of ergodic capacity are analyzed on the different conditions and scenarios. The distribution of the mutual information is investigated on the doubly correlated Rician fading channel, and the outage capacity is analyzed. The capacity bounds of MIMO-OFDM system are investigated in correlated fading channels. The bounds of frequency-selective Rayleigh fading are deduced and proved. The outage capacity of MIMO-OFDM system is analyzed. Assuming the mutual information is approximately Gaussian distributed, the variance is derived for the correlated frequency-selective Rician channels. With the channel model, we demonstrate that the approximation is accurate over a range of system parameters.Finally, MIMO information theory with feedback is discussed. The capacity of spatial-temprol linear precoder is compared with that of the traditional MIMO system. Aiming at the multimedia application, a virtual MIMO technology is proposed for the precoding system. With different targets, the power distribution strategies are given for different QoS.