Analysis of Deep Learning-based MIMO Detectors

Abstract

Multiple-Input Multiple-Output (MIMO) communication systems have become a fundamental technology in modern wireless networks due to their ability to enhance data rates and system capacity. However, traditional MIMO detection algorithms face significant challenges, including increasing complexity and performance degradation with growing system dimensions. Deep learning has shown great promise in various domains in recent years, leading researchers to explore its potential in addressing MIMO detection capability. This paper provides a comprehensive overview of deep-learning-based MIMO detection techniques, presenting an extensive literature review and taxonomy of approaches. We conduct a comparative analysis with conventional techniques and evaluate the detection performance of deep learning-based approaches. The paper also compares the required number of FLOPs operations to identify the potential of each detection method.