Performance Enhancement of Scalable Cell-free Massive MIMO Systems Using Dynamic Cooperation Cluster Optimization Concept

Abstract

Downlink cell-free massive MIMO is an emerging wireless communication system that aims to overcome the limitations of traditional cellular networks. In contrast, a cell-free massive MIMO network uses a large number of distributed access points (APs) or remote radio heads to provide coverage, without the concept of distinct cells. This allows for more uniform signal coverage and helps eliminate cell-edge problems. However, the difficulty lies in realizing the benefits of cell-free operation in a practical manner. These challenges are Efficient coordination, channel Estimation, Scalability, Hardware complexity and so on. In response to this challenge, this thesis presents a dynamic cooperation cluster concept for scalable Cell-Free Massive MIMO systems. Leveraging the dynamic cooperation cluster concept, the Gradient Descent based Minimum Mean Square Error approach introduces an algorithm for joint initial access, and cluster formation that has been rigorously proven to be scalable. The gradient descent optimization technique is employed to progressively refine the MMSE precoding matrix. Additionally, it adopts precoding techniques to ensure scalability. Furtherly enhances the spectral efficiency, signal-to-interference-plus-noise ratio and bit error rate (BER) compared with Partial MMSE (Minimum mean square Error), LMMSE (Linear minimum mean square Error), RZF (Regularized Zero Forcing) precoding. The performance gap between the proposed method and the other techniques increases as the SNR, N, and number of UEs increases. This implies that the proposed method is particularly effective in scenarios with high SNR, number of UEs. This is achieved by calculating the gradient of the mean squared error (MSE) with respect to the precoding matrix PMMSE, and then using this gradient information to iteratively update the precoding matrix. The goal of this approach is to minimize the MSE between the intended signal and the received signal, and improves the performance of desired system.