Transmit Beamforming with Reduced Channel Information in OFDM Based Wireless Systems

In this paper, we consider transmit beamforming that works with reduced channel state information in OFDM based multiple-input single-output (MISO) wireless systems. The proposed scheme significantly reduces the amount of feedback signaling burden by generating beamforming weights using information on the previous beamforming weights and channel correlation. The feedback signaling overhead is further reduced with the use of clustering and interpolation techniques. Simulation results show that the proposed scheme outperforms conventional beamforming techniques, while using the same amount of feedback signaling overhead. encoding the codeword with the use of previous ones [7]. The amount of feedback signaling information can further be reduced by employing a so-called clustering technique that combines adjacent subcarriers into a cluster and uses a same beamforming vector for these subcarriers [8]. Performance degradation due to the use of a large cluster size can be alleviated by interpolating the beamforming vectors in the frequency domain [8]. However, it may require additional phase information. In this paper, we consider the reduction of the feedback signaling burden for the transmit beamforming by exploiting the channel correlation in the frequency domain. The beamforming vectors can be generated in a sequential manner using the correlation characteristics between the adjacent subcarriers. For instance, when the adjacent subcarriers are correlated to each other, the beamforming vector of the current subcarrier can be represented in terms of that of the adjacent subcarrier and a vector corresponding to the difference between the two adjacent subcarriers. Since the vector corresponding to the difference will have a span smaller than the original one, the resulting quantization error will be smaller than that of the original one when the same amount of feedback signaling is considered. Further reduction can be achieved by using interpolation techniques with the use of a phase-dependent codebook.