Main Article Content
Performance Evaluation of Full Array, Sub-Array and Overlapped-Sub-Array Hybrid BeamForming Architectures for Massive MIMO Systems
Abstract
The technological advancement in wireless communication, promises high data rate for end users. This has led to the possibility of smart cities, inter connected vehicles, and virtual reality applications. One of the recent technologies in wireless communication is massive MIMO where large number of antennas are deployed at the transmitter or receiver. This is possible due to the use of mmWave in wireless communication. With massive MIMO, beamforming technique can be employed in the communication system. Beamforming is the ability of communication system to direct power to the intended users and to cancel power at non-intended users and thus significantly improving communication system performance. Digital beamforming was initially used in MIMO systems. However, for massive MIMO systems it leads to high power consumption due to large number of dedicated radio frequency chain in each antenna. To address this challenge, hybrid beamforming techniques were introduced. There are three architectures for hybrid beamforming: Fully array architecture (FAA), overlapped sub-array architecture (OSA) and sub-array architecture (SAA). This paper has analysed three performance parameters of the mentioned hybrid beamforming architectures. The simulation results show that, FAA architecture has high performance in outage probability and spectral efficiency. However, its energy efficiency is lower compared to OSA and SAA. Specifically, SAA has the highest energy efficiency in comparison to FAA and OSA. It can also be observed that, with only 25% increased number of elements in OSA, the energy efficiency can be slightly lower compared to SAA, while achieving appreciable spectral efficiency performance with respect to FAA. Additionally, this work has derived an outage probability expression, which has not been covered in most of the studies. This study gives an insight of selecting the best architecture based on the performance requirement.