Study On Limited Feedback Beamforming Algorithm Over The Codebook-Based MIMO System

The new generation of wireless communication system is required to provide high data rate as well as service quality. However, wireless spectrum resources are limited. And after years of exploitation, there is no more left. So, the development of spatial resources attracts people's attention. MIMO technology can increase system capacity without extra bandwidth. It becomes one of the most popular technologies in the next generation of wireless communication system.MIMO systems can achieve significant diversity and array gains by using transmit beamforming and receive combining techniques. In the absence of full channel knowledge at the transmitter, the transmit beamforming vector can be quantized at the receiver and sent to the transmitter using a low-rate feedback channel. The best beamforming vector is the right singular vectors corresponding to the largest singular value of channel matrix. Only feedbacks this vector rather than the whole channel matrix can still provide good system performance. This method is called limited feedback.The traditional VQ (Vector Quantization) algorithms and Grassmannian Line Packing have only considered the independent and identically distributed Rayleigh flat fading channel. But the reality channel often exist in the spatial correlation or time correlation, even exist temporal-spatial correlation. Most of the codebook design algorithm only considers one kind of the channel correlation. Therefore, we need to study an adaptive codebook design algorithm that can apply to various channels.In this paper, we propose an adaptive codebook design algorithm and codeword search algorithm which include many child books. Each child book consists of three components: characteristic direction, local perturbation and global perturbation. There is one-one map from each characteristic direction to each child codebook. When search codeword, if the number of codeword is in the area of characteristic direction or local perturbation, then next time still use the current child codebook, otherwise, next time use the child codebook corresponding to the number of the current codeword. Therefore only the number of codeword needs to feedback. And codebook update only when the spatial correlation of channel is changed.Computer simulation results show that our adaptive codebook design algorithm has good performance in all kinds of channel correlations.