High Efficiency Beam Alignment Based on Multi-Modal Beam Patterns for Massive MIMO Antenna Systems

Multiple-input multiple-output (MIMO) is considered as the key technology for 5G wireless communications networks. In the millimeter-wave (mmWave) frequency bands, directional beamforming is essential to the establishment of an efficient radio link for signal transmission. However, in wireless mobile environments, the effective beam direction is generally unavailable in the initial access phase. Hence, beam alignment (BA) is crucial for link establishment in initial access. In this work, we propose an efficient BA technique, referred to as the multi-modal pattern BA scheme, to improve the performance and efficiency of initial access in massive MIMO systems. Unlike the conventional sequential search scheme using single-modal patterns, the transmission of synchronization signals in the proposed scheme is based on the use of multi-modal patterns, each of which specifies multiple main lobes with distinguishable directions. Incomparison with the conventional scheme, our scheme can greatly enhance the achieved angular resolution after BA under the use of a fixed number of synchronization signals. As a result, the signal-to-noise ratio (SNR) of the established link is significantly improved. By contrast, if the same angular resolution is desired, the proposed scheme can greatly reduce the number of required synchronization signals to conserve the radio resources, including transmission power and bandwidth, and to decrease the searching delay in the BA procedure. In addition, considering multi-modal pattern construction for the transmission using a single radio-frequency (RF) chain, we propose an effective approach to generate the desired beamforming vectors fulfilling the essential constant amplitude constraint.