Performance analysis, pilot designs, and algorithms for massive MIMO system

There is continuing growth in demand for data rate. Data rate of tens of megabits per second should be supported for each of tens of thousands of users, and this brings its own technical challenges. One of the solutions recently proposed in the literature is the use of many antenna elements, and such a system is called massive Multiple-Input Multiple-Output (MIMO) system. Having closed-form achievable rate is very useful from both theoretical and practical perspectives. In this dissertation we develop closed form achievable rate for downlink (DL) of massive MIMO under pilot contamination problem with and without DL pilot transmission. Also we develop performance bounds on achievable rate in uplink (UL) of massive MIMO. Then we propose a new channel estimator for a multi-cell massive MIMO scenario with performance close to the ideal Minimum Mean Square Error (MMSE) channel estimator but less knowledge of long term statistics of channels. Also we develop power allocation for massive MIMO based on maximizing sum-rate.;Then we tackle one of the main challenges of massive MIMO, which is working in Frequency Division Duplex (FDD) mode. Due to large pilot overhead, usually in the literature it is assumed massive MIMO operates in Time Division Duplex (TDD) mode. But, many practical systems work in FDD mode over different frequency bands. In this thesis we propose new transmission/transceiver architectures for massive MIMO called rotated FDD (RFDD) which let the massive MIMO work in non-contiguous bands. In addition, we develop and analyze rate adaptation mechanism between UL and DL when operating in non-contiguous bands. This UL and DL rate adaptation has been viewed as a fundamental limitation for systems with FDD spectrum assignment, and our approach offers an efficient solution to this long standing problem.;Next, we propose novel pilot designs and a compensation scheme for millimeter-wave massive MIMO systems with severe RF distortions including phase noise and In-phase Quadrature (IQ) amplitude and phase imbalance at both transmitter and receiver sides.