Spectral Efficiency Evaluation of Downlink Cell Free Massive MIMO Using Hybrid Precoder

Abstract

The rapid advancement of communication technologies is connecting the world faster than ever before. However, this has resulted in a scarcity of frequency spectrum as the available communication bandwidth is limited. To maximize data transmission rates while using a fixed bandwidth, a mechanism is needed to improve spectral efficiency. Distributed massive multiple input multiple output (MIMO) systems involve deploying a large number of transmitting antennas around user equipment (UEs) to achieve this goal. However, in cellular massive MIMO, signal interference around cell edges can present a problem. To address interference issues around cell edges in cellular massive MIMO, cell-free massive MIMO was introduced, which incorporates both massive MIMO and distributed MIMO principles. Cell-free massive MIMO is advantageous as it places users closer to access points (APs), enabling higher spectral efficiency. While cell-free massive MIMO improves spectral efficiency, signal processing techniques such as precoding are needed to maximize it further. Precoding is a transmitter signal processing method that maximizes the received signal for specific receivers while minimizing interference for all other receivers. This thesis work proposes a better performing precoding algorithm for cell-free massive MIMO based on partial regularized zero-forcing (PRZF) and the marine predator algorithm (MPA). Compared to the PRZF precoder, the proposed algorithm has resulted in an average spectral efficiency improvement of 11.8% for 20 users and 13.65% for 100 users within the coverage area, assuming that the network is equipped with 1500 APs, each having a single antenna. Additionally, it has improved the sum spectral efficiency by 13.6% for 60 users within the coverage area, given that the network is equipped with 1500 APs, each having a single antenna. Furthermore, the computational complexity of the proposed method is relatively comparable to that of existing precoders.