Soraya Norma Mustika, Muladi, Triyanna Widiyaningtyas
Due to the constraints of multiple inputs, multiple outputs (MIMO) numerous researchers have created an extensive MIMO system. The uplink MIMO systems pose one of the most significant challenges in large linear detector MIMO systems. MIMO systems must do important matrix inversions to estimate the transmitted data. By enabling channel estimation based on sparse Bayesian learning techniques, we access data using MATLAB. This research implements singular value decomposition (SVD) for zero forcing (ZF), minimum mean square error (MMSE), and two standard linear detection techniques. The SVD, given rotation, and Golub Reinsch algorithms are applied whenever some user joins or leaves the base station(BS). If we intend to minimize error, we must use the Reinsch algorithm. The complexity is more significant when the Reinsch algorithm is repetitive. This study attempts to develop a substitute for the MIMO uplink to make it easier to determine how valuable the current channel constitutes a conflict of interest. As a result, we are attempting to create a novel method for approximating transmitted data while decreasing the complexity of the inverse matrix's enormous size. We represent our method's cumulative distribution function and less complexity compared to various alternative approaches such as conventional MMSE and ZF Newton iteration, Neumann Series(NS), and Gauss-Seidel. This is showed in error rates of 1% being considered and less complexity. This research followed ethics rules, was unique to the authors, and was not previously published. © 2025 IEEE.
University of State Malang, Dept of Electrical Engineering, East Java, Malang, Indonesia; University of State Malang, Faculty of Applied Science Engineering, East Java, Malang, Indonesia
