MS Thesis Defense
On MIMO Channel Modeling and Capacity
Using the Channel Correlation Matrix
Anush Chandrasekaran
1:00pm Wednesday, 18 July 2012, ITE 325b
Communication systems have always been affected by multipath propagation that causes a delay and distortion in receiving the signal, with a different delay for each path. Multiple-Input Multiple-Output (MIMO) communication systems were developed to combat this problem and use multipath propagation to their benefit. A MIMO communication system contains M transmitter antennas and N receiver antennas that are used to improve either the robustness of transmission or the throughput.
We assume an exponential channel correlation matrix R model for the MIMO channel with J = M = N and use it to compute the channel H-matrix, the receiver (RRx) and transmitter (RTx) correlation matrices, and the ergodic MIMO channel capacity (CH). We propose two algorithms to obtain RRx and RTx from R, which have been used to estimate/bound CH. We investigate and compare three ergodic MIMO channel capacity estimation/bound methods for our MIMO channel model in this thesis. The first two existing estimation/bound methods use the Kronecker model and an RRx-based bound, respectively. The third method is a novel method we propose and study to estimate the ergodic MIMO channel capacity using specific eigenvalues of RRx. The behavior of the eigenvalues of R and RRx are analyzed to identify the eigenvalues that can be used in this method. This method achieves less relative-error compared to the RRx-based bound. It is better than the Kronecker model for specific values of J and the correlation parameter r.
Committee: Drs. Joel M. Morris (Chair), E. F. Charles LaBerge, Mohamed Younis and Tinoosh Mohsenin