Multiuser MIMO Systems Channel Estimation with Gaussian Distributed Data Symbols

Massive MIMO technology uses large arrays of antennas at the base station, often numbering in the low tens to hundreds, to greatly increase capacity, diversity, and spectral efficiency. Nevertheless, these benefits are contingent upon the receiver's ability to precisely get channel state information (CSI) for both downlink as well as uplink channels. Pilot sequences are usually supplied and used by the recipient to identify the CSI. The more transmit antennas used in the massive MIMO system, the more pilot sequences are required. This requirement for lengthy pilot sequences can lead to lower spectral efficiency and introduces an issue known as pilot contamination, which occurs when pilot sequences are reused in neighboring cells.The effectiveness of semi-blind estimation approach, the Cramer-Rao bounds (CRB’s) for this method of estimation of channel have been calculated for both stochastic and deterministic (Gaussian) data symbol models. By studying how the CRBs change as the number of antennas at the base station increases, insights into the asymptotic performance of massive MIMO systems are gained. Numerical analyses highlight the advantages of the semi-blind estimation technique, particularly in terms of reduced mean squared error. Additionally, as number of antennas grows, The Expectation Maximization (EM) technique performs similarly to the maximum likelihood estimator with genie assistance, assuming that the data symbols are known.