dc.description.abstract |
Multiple-input multiple-output (MIMO) techniques can be used to achieve
diversity gain, multiplexing gain and/or array gain. Particularly, diversity cod-
ing techniques (e.g. Space-Time (ST), Space-Frequency (SF) coding) have re-
ceived tremendous attention as e ective means of achieving spatial diversity gain
in MIMO systems. However, in the presence of spatial correlations the diversity
gain of ST/SF coding diminishes. Beamforming techniques can be used to achieve
array gains in MIMO systems. Hence, in correlated channels beamforming tech-
niques can be combined with ST/SF coding to further improve the performance.
In this thesis, we develop beamforming techniques relying on statistical channel
state information at the transmitter (CSIT) for space-time (ST) / space-frequency
(SF) coded MIMO systems to minimize the pair-wise error probability. We pro-
pose beamforming techniques for SF coded MIMO-OFDM systems in correlated
frequency-selective Rician fading channels. We propose two novel beamforming
techniques for this channel model.
Furthermore, distributed beamforming techniques are developed for correlated
Rayleigh
at-fading channels, relying on full-instantaneous CSIT as well as sta-
tistical CSIT. Moreover, we extend these techniques for SF coded MIMO-OFDM
relay networks in correlated Rician fading channels and propose optimal beam-
forming techniques relying on full-instantaneous CSIT. Also, suboptimal beam-
forming techniques relying on statistical CSIT are developed. The variation of
error performance is thoroughly investigated and the simulation results con rm
that all the proposed beamforming techniques achieve signi cant performance ad-
vantages over MIMO systems using ST or SF coding only. |
|