Cascaded and Separate Channel Estimation based on CNN for RIS-MIMO Systems

Abstract

With the dramatic increase in mobile users and wireless devices accessing the network, the performance of 5G wireless communication systems is severely challenged. Reconfigurable Intelligent Surface (RIS) has received much attention as one of the promising technologies for 6G due to its ease of deployment, low power consumption, and low price. This study aims to improve accuracy, reliability, and the capacity to estimate channel characteristics between transmitter and receiver. However, this is practically challenging for the following reasons. Due to the lack of active components for baseband signal processing, low-cost passive RIS elements can only reflect incident signals but without the capability to transmit/receive pilot signals for channel estimation as active transceivers in conventional wireless communication systems. This study presents different channel estimation methods for RIS-MIMO systems that use deep learning techniques.

Keywords

• RIS-MIMO
• phase shift
• channel estimation
• deep learning
• beamforming