Downes, Garfield Orlando2013-11-072013-11-0720062006Source: Masters Abstracts International, Volume: 45-02, page: 0987.http://hdl.handle.net/10393/27240http://dx.doi.org/10.20381/ruor-18607Wireless communication using multiple-input multiple-output (MIMO) technology enables increased spectral efficiency. The successful deployment of commercial MIMO systems requires a strong understanding of the channel conditions to achieve desired BER performance. A simple channel model that is frequently used in MIMO research assumes i.i.d. complex Gaussian distributed channel gains to describe the channel links. However, this model ignores common propagation effects present in outdoor environments. On the other hand, the wireless channel is very complex to describe using closed-form expressions. The objective of the work presented in this thesis is to parameterise the wireless channel using metrics and use these to predict the bit-error rate (BER) performance for different simulated MIMO techniques. The data measurements were obtained in Ottawa, Ontario at approximately 2 GHz, using the CRC MIMO testbed. Communications systems employing spatial multiplexing or space-time block coding (STBC) are mainly designed for operation under ideal channel characteristics. We consider (1) Vertical-Bell Labs Layered Space-Time (V-BLAST) architecture and (2) a QPSK-based quasi-orthogonal STBC. The measured data exhibit different location-specific characteristics. In some regions, the spatial correlation is low resulting in high diversity and high available mutual information. In others, spatial correlation is higher. Of all the metrics considered, we have identified six that provide useful information to describe the wireless channel. These include effective channel rank, correlation and diversity. The impact of signal power has also been investigated. It has been found that the performance spatial multiplexing is dominantly dependent on the smallest eigenvalue of the spatial correlation matrix. It is not, however, as useful in predicting the performance of STBC. Spatial correlations appear to improve the BER performance of STBC by improving the effective average SNR. It is important to note this trend is opposite to that found in spatial multiplexing. The largest eigenvalue also appears to have an impact on STBC performance. Also, it is shown that mutual information is sufficient enough to describe channel behaviour. These metrics give insight into the dominant propagation mechanisms in the environment and demonstrate the effectiveness of this kind of measurements in characterising the behaviour of the mobile radio channel.119 p.enEngineering, Electronics and Electrical.Thesis