An Improved Codebook Training Procedure using Compressed Sensing in mmWave Hybrid Beamforming

摘要

Millimeter Wave (mmWave) communication with the advantage of high spectrum availability is a key technology to offer several gigabits-per-second data rates for future 5G and beyond applications. Limited power availability and smaller coverage area are some critical issues in it due to its high attenuation and atmospheric absorption. Beamforming has become a skillful technique to keep track of the angle of arrival (AoA) and angle of departure (AoD) by directing the steering vector along with these phases with the least beam-width. Providing an optimal power allocation towards a user’s location through beamforming has become an efficient technique to direct beam along the angle of departure (AoD) of the transmitted wave with the smallest beam-width. In a multi-input and multi-output (MIMO) channel, the channel state information (CSI) of a large dimension array leads to a considerable feedback overhead. The sparse nature of the MIMO channel has been exploited by researchers using compressed sensing, where, the resulting precoder indices can be sent back to the transmitter in order to reduce the feedback overhead. The precoder and combiner are updated with their estimated AoA and AoD during the training period. However, the training time goes enormously high as the number of antennas become too large. To overcome such limitation, an improved codebook training procedure (CTP) incorporated with hybrid beamforming (HBF) has been suggested in this research. The proposed scheme will enable the highly directive beam to track the dynamic location of multiple mobile users more efficiently and accurately.