Amalia Eka Rakhmania, Hudiono, Umi Anis Ro’isatin, Nurul Hidayati
—Massive Multiple Input Multiple Output (MIMO) has emerged as a crucial technology in 5G and future 6G networks, offering unprecedented improvements in capacity, energy efficiency, and spectral efficiency. A key challenge for Massive MIMO systems is accurate and efficient channel estimation, which significantly impacts system performance. Traditional channel estimation methods such as Least Squares (LS) and Minimum Mean Square Error (MMSE) have been widely employed, but their limitations, particularly in complex and dynamic environments, have led to the exploration of more sophisticated approaches, including machine learning (ML)-based techniques. This review aims to compare traditional channel estimation methods with modern machine learning-based techniques in Massive MIMO systems, providing insights into their performance, computational complexity, and scalability. Furthermore, this paper outlines potential future research directions, emphasizing the integration of machine learning, optimization techniques, and hardware-friendly designs for enhanced performance. © 2025 Scientific Association for Infocommunications. All rights reserved.
Department of Electrical Engineering, State Polytechnic of Malang, Indonesia