Cluster Based MIMO Radio Channel Modeling
MetadataShow full item record
Modeling of wireless channels is perhaps the most complex task in implementing a wireless communication system because wireless channels change over time, vary with frequency and space. That is, the transmit signal with limited power will be distorted at the receiver over time, frequency, and space simultaneously. In order to better understand the mechanism behind it, we need to study this physical phenomenon. In this dissertation, we model the MIMO-OFDM channels in a multi-cluster radio wave propagation environment. In other words, we use a single distant scattering cluster to study the second-order statistics of the channel and build MIMO-OFDM simulation channels in accordance with the correlation properties of the channel. In this approach, each distant scattering cluster contributes a portion to the Doppler spectrum and corresponds to a state space channel model. Two types of random scattering clusters are proposed, namely the Cauchy-Rayleigh cluster and Rayleigh cluster, to model the power spectrum of the AOA, AOD and the delay power spectrum of the TOA. According to the trigonometric relationship among the transmitter (receiver), cluster center, and scatterers, we approximately map the spatial location of the scattering objects to as a Cauchy or Gaussian angular power distribution function. These two distribution functions made the secondorder statistics of the channel integrable. Hence, we obtain analytical solutions about the channel spatio-temporal correlation function. In order to separate the dynamics and antenna spacing of the channel spatiotemporal correlation functions into disjoint parts, we propose a linear transformation. This allows us to model the channel dynamics caused by the Cauchy-Rayleigh cluster as an AR(1) model, and the dynamics caused by the Rayleigh cluster approximately as an AR(3) model. The antenna spacing thus becomes part of the initial point of movement. The conversion from the channel temporal dynamics to the AR(1) model is very straightforward, while for the AR(3) model, some methods are required. We convert the AR(1) and AR(3) models into a state space form to create the SISO channel model. That is, each AR model corresponds to a radio channel and is associated with a distant scattering cluster, which is represented by a state space SISO channel model block. A MIMO channel model is then constructed by connecting multiple SISO channel blocks in parallel, in which a coloring matrix is used to adjust the channel spatial correlation between the SISO blocks. A MIMOOFDM channel model is obtained in the same manner. This time, however, a coloring matrix described by the Kronecker correlation matrix is used to adjust the channel spatial correlation within each MIMO channel block and the channel spectral correlation between the MIMO blocks. This approach has three advantages: simple, the entire Doppler power spectrum can be formed from multiple uncorrelated distant scattering clusters, and the channels contributed by these clusters can be obtained by the summing of the individual channels. In this way, we can reassemble the radio wave propagation environment in a simulated manner.