Dictionary Based Channel Estimation and Precoding Techniques for Massive MIMO Wireless Communication Systems
Abstract
Massive multiple-input multiple-output (MIMO) systems constitute a promising enabling technique for 5G/6G cellular networks as a benefit of their substantial spatial multiplexing gain in both time division duplex (TDD) and frequency division duplex (FDD) scenarios. In particular, the millimeter-wave (mmWave) band, has emerged as one of the leading candidates for spectrum exploitation to address the impending spectrum scarcity and to facilitate high-speed data delivery. To achieve these gains, knowledge of channel state information (CSI) is essential at the base station (BS) to implement transmit precoding, which enhances spectral efficiency. In a TDD system, by exploiting the channel reciprocity property, the BS can estimate downlink CSI from the sounding on the uplink channel. But in a FDD system due to the absence of channel reciprocity, it is required to compress the CSI at the user equipment (UE) and feed it back to the
BS. Feeding back the accurate CSI becomes more challenging with the increased number of antennas, subcarriers, and UEs. The compression of high-dimensional CSI is essential for reducing the CSI feedback. Compressive sensing (CS) is a promising approach for reducing CSI feedback compared to the quantization codebook scheme for the same bit
error rate performance. With CS, the high dimensional sparse channel can be compressed into a low dimensional channel. In this thesis, the primary focus is on compressing and estimating the CSI, as well as utilizing the estimated CSI for precoding in FDD systems. The performance analysis presented in this thesis explores various aspects of massive MIMO systems, including narrowband and wideband systems, as well as the roles of reconfigurable intelligent surfaces (RIS) and reconfigurable holographic surfaces (RHS) in the mmWave band.