With an ever-increasing demand for higher wireless throughput, there has been growing interest in massive multiple-input multiple-output (MIMO) as a key technology for future wireless networks. This dissertation addresses some of the key aspects of this technology that include: 1. precoding, power optimization, and access point (AP) location design in cell-free massive MIMO systems with distributed APs; 2. semi-blind channel estimation in massive MIMO systems.

Cell-free massive MIMO is a special deployment of massive MIMO systems with a large number of distributed low-cost low-power single antenna APs serving a much smaller number of users. The cell-free system is not partitioned into cells and each user is served by all APs simultaneously. The downlink capacity lower bounds for conjugate beamforming and zero-forcing precoders in cell-free systems are derived in this dissertation. To further increase the achievable throughput, max-min power optimization algorithms are formulated, and low complexity max-min power allocation algorithms are developed. We also introduce a technique that employs L1-norm sparsity penalty in the max-min power optimization for conjugate beamforming that helps us decrease the number of APs that serve a user in a practical system. The uplink capacity lower bounds for minimum mean squared error (MMSE) and large scale fading decoding receivers in cell-free systems are provided. A deterministic approximation for signal-to-interference-plus-noise ratio of MMSE receiver is obtained with an unlimited number of APs and user devices.

Next, AP location design problem is investigated to maximize the sum-throughput and the minimum-throughput in uplink transmission of cell-free systems with an arbitrary user distribution. Utilizing compressed sensing techniques, the AP placement problems are formulated as convex optimization problems. An AP location design algorithm is also presented in an alternative small-cell system in which each user is served by only one AP.

Finally, semi-blind channel estimation for multiuser massive MIMO systems is investigated. Multiple semi-blind channel estimation techniques based on the expectation-maximization algorithm are developed by considering different priors on data symbols. Cramer Rao Bounds (CRBs) for semi-blind channel estimation are derived for deterministic and stochastic (Gaussian) data symbol models to give us an analytical understanding of the semi-blind scheme’s performance. To get insight into the behavior of a massive MIMO system, the asymptotic behavior of the CRBs as the number of antennas at the base station grows is analyzed.