Efficient codebook-based MIMO beamforming for millimeter-wave WLANs

Beamforming is a critical and natural solution for 60 GHz band wireless communications due to the increased signal attenuation. In this paper we propose an efficient codebook-based MIMO beamforming training scheme for estimating the antenna weight vectors (AWVs) in millimeter-wave wireless local area networks (WLANs). We adopt the discrete Fourier transform (DFT)-based codebooks. We propose a simple and efficient iterative AWVs training algorithm that uses only one AWV feedback to obtain the best transmit (TX) and receive (RX) AWVs pair. Instead of initial selection of multiple TX and RX AWVs in the IEEE 802.11ad draft standard, we only select the best TX AWV and RX AWV for subsequent refinement. The beam pattern search for refinement is limited only to the region corresponding to the best TX AWV and RX AWV from initial selection and their angular rotated adjacent beams. To verify the performance we compare our proposed scheme with the one specified in the IEEE 802.11ad in terms of the beam gains and show that our proposed scheme outperforms the one in the IEEE 802.11ad even using the fewer beamforming training time.