A Study On The Model Of MIMO Channels

The multiple-in-multiple-out(MIMO) technology has been regarded as one of the keytechnologies of the next generation mobile communication system (4G), because it can extend thechannel capacity of wireless communication system without increasing the channel bandwidth.MIMO channel model is necessary both at the phase of research of MIMO technology and thecommercial application of the MIMO wireless communication system. Therefore, the modeling ofMIMO channel play a very important role in relevant research. There are two kinds of MIMOchannel model, deterministic model and stochastic model. The latter includes geometrically model,parametric model and correlation model. The geometrically stochastic model is widely investigatedbecause of the advantages such as simple structure, clear physical concept, strong adaptation ability,and convenience to simulate on a computer.Refered to the micro-cell ellipse model of smart antenna system, a new micro-cell MIMOchannel model, ellipse model, is presented in this thesis, based on the sufficient analysis of thecharacteristics of micro-cell MIMO channel. The realistic micro-cell MIMO scattering environmentcan be simulated with the assumption that the scatterers are stochastically distributed on the ellipseand that the Base-Station(BS) and Mobile-Station(MS) are at the foci. The distribution of scattererson the ellipse is assumed as a uniform distribution of angle-of-arrival(AOA) on MS. Compared withthe two-ring model, the line-of-sight(LOS) and non-line-of-sight(NLOS) are both taken intoconsideration in this model. By use of this model, the computational consuming is reduced. Basedon the ellipse MIMO channel model, the effects of antenna space, Rice coefficient, moving speed ofMS and delay time to space-time correlation coefficients are studied by simulation. The resultsshow that enlarging the antenna space or accelerating the MS moving speed can decrease thecorrelation between antenna links. This conclusion is similar to that from other theories and theagreement indirectly verifies the applicability of this model.Combined with the characteristics of the micro-cell MIMO channel, a new indoor MIMOchannel model is presented, in which the scatterers are uniformly distributed on an ellipse, the MSis fixed and the ellipse size is smaller for the indoor environment. Similar to the outdoor model, thegains of the link can be derived. By simulating, the effects of spaces between antennas and Ricecoefficient on correlation coefficients are analyzed in detail. The simulation results are given toshow the consistency with the theoretical estimation.In this thesis, based on the outdoor micro-cell MIMO channel model, the channel capacity ofrealistic micro-cell MIMO channel is simulated using Monte Carlo method. The results show thatthe increasing of antennas can enlarge the channel capacity. With the condition that numbers oftransmitting and receiving antennas are the same, the capacity can still increase linearly with thenumber of antennas, However, if those numbers are different, the capacity is determined by thesmaller number of the antennas. The results also show that the channel capacity decreasinglyincreases with the antenna spacing and decreases with the coefficient K. The results using proposedmodel possess the similar conclusion to those of other model. The similarity implies the validity ofthe model presented in this thesis.