Hybrid Beamforming for Coded MU Massive MIMO-OFDM System

Abstract

With the ascent in users of Internet of Things and mobile data, an enormous growth has been noticed in the usage of data in recent years and if this trend remains unchanged, it will increase as multiples of zettabytes in the near future. So, for the accomplishment of the requirements of the future generations, the (Multiuser Massive Multiple Input Multiple Output-Orthogonal Frequency Division Multiplexing) MU Massive MIMO-OFDM system is being evolved as an emerging technology. Therefore, for assessing its performance, its comparative analysis in terms of error and achievability rates has been performed with the earlier MIMO-OFDM techniques and it has been found that this scheme outperforms the earlier ones with improved rates. As for the downlink transmission case, to achieve 0.01 (Bit Error Rate) BER, (Signal to Noise Ratio) SNR required for 4x4 MIMO-OFDM system is 9dB, 8X4 MU MIMO-OFDM system is 4dB and 128x4 MU Massive MIMO-OFDM system is -3.8dB. Further for actualizing this system in practical scenarios, Phase Zero Forcing (PZF) hybrid beamforming scheme has been introduced in the system and is compared with the earlier Zero Forcing (ZF) scheme in terms of BER, achievability rate and computation complexity. From the obtained observations, it has been found that this PZF scheme performs almost equivalent to ZF scheme with small loss in system’s performance which is reimbursed in terms of reduced computational complexity. This reduction in complexity for the massive MIMO system has been measured by plotting the graph for time consumption as a function of service antennas which shows that for instance of 600 service antennas, the computational time required for ZF scheme is 640 seconds whereas for PZF scheme is 250 seconds. In addition to above-mentioned analysis, one more category of comparison between the distinct FEC codes and un-coded system has been assessed and evaluated. After examining these comparisons, it has been observed that these systems surpass the performance of un-coded system but at the cost of raised computational complexity.