Deep Learning-based Channel Estimation in 5G MIMO-OFDM Systems

Abstract

The primary air interface of fifth-generation (5G) wireless communication is multi-input multi-output orthogonal frequency division multiplexing (MIMO-OFDM). In addition, channel estimation holds a major role in achieving effective system performance in 5G wireless networks. Importantly, deep learning (DL) techniques have the ability to improve system performance and reliability and to reduce the computational complexity of 5G communication systems. Hence, DL techniques have been applied in MIMO-OFDM systems to improve the channel estimation quality and reduce the bit error rate compared to traditional channel estimation methods. This paper offers a comprehensive survey of DL-based channel estimation in MIMO-OFDM receivers in terms of DL algorithms, 5G channel models, and error performance. In conclusion, the application of DL techniques in MIMO-OFDM shows a significant potential for performance enhancement in 5G networks.