LAUSR

In a multi-user multiple-input multiple-output (MU-MIMO) communication system, multiple RF (radio frequency) chains are generally connected corresponding to the antenna elements. However, the system complexity, hardware cost, and energy consumption are challenging while using many RF chains. In addition, an information-centric Internet of Things (IoT) will be significant to guarantee fast service demand for radio access in the fifth generation (5G) mmWave wireless systems. In this paper, a new beam training method based on a single RF chain is proposed to overcome these challenges by using the downlink–uplink and downlink–downlink beam training techniques to establish the elite users’ subset group which can be trained in a single training time slot. Additionally, the users’ subset mechanism is implemented by selecting an arbitrary function to guarantee the convergence. Moreover, the proposed techniques are compared with the conventional full search method regarding searching time, cost, and complexity. Different user subset functions are modeled, evaluated, and compared with the existing ones with respect to effective channel gain and average system capacity. The simulation results show that the proposed method can achieve a significant system performance, and the single RF chain is serviceable for a large number of users when connected with a maximum of 64 antenna elements at the BS. The proposed method can also be implemented in a massive MU-MIMO system by using multiple RF chains in which each RF chain is connected to a group of 64 antenna elements. This work is applicable in 5G mmWave communications, which require robust infrastructure to meet the future IoT service demand in 5G hot scenarios.