Abstract:
In highly spatially-correlated channels the diversity gain obtained by diversity coding schemes (such as Space-frequency (SF) coding or Space-Time (ST) coding) is considerably low. Combining beamforming with SF/ST coding help to improve the system performance in spatially correlated channels. In this paper, we propose two beamforming technique for SF coded MIMO-OFDM systems in a frequence-selective Rician-fading channel to minimize the pair-wise error probability (PEP). The proposed beamforming techniques are based on the channel mean and spatial correlation statistics. In Technique, 1. all the eigenmodes of the channel spatial correlation and mean matrices taken in to consideration, where in Technique 2. Only the strongest egimodes of the channel spatial correlation and mean matrices is considered. Simulation results demonstrate that Technique 2 offers a substantial performance improvement over SF coding without beamforming in highly spatially-correlated channels. Furthermore, considerable performance improvements can be obtained using Technique 1 overSF coding only in low correlated channels with a high Ricean K factor.
