Fast beam alignment algorithm for multi-user mmWave communications
The training overhead in pursuing pencil beams is commonly heavy for millimetre-wave (mmWave) communications, particularly in multi-user (MU) scenarios with conventional sequential-training methods. To reduce the training overhead, a parallel beam alignment algorithm is proposed for the MU mmWave system with a hybrid structure. The proposed algorithm jointly exploits parallel beam alignment among different user-subsets and hierarchical beam training for each user-subset. Consequently, the training overhead is significantly reduced to for the mmWave systems with transmit antennas. Numerical results verify that the proposed algorithm outperforms traditional alternatives in MU scenarios.