Performance Analysis of Multi-User Massive MIMO Systems Subject to Composite Shadowing-Fading Environment

Multiple single antenna terminals transmit simultaneously to an array of hundreds of antennas in the uplink of a multi-user massive multiple-inputmultiple-output (MIMO) system. Under Rayleigh fading and lognormal shadowing, the expression for the probability density function (PDF) for signal-to-interference-plus-noise-ratio (SINR) does not exist in a closed-form, which is required to calculate the outage probability of a user. In this thesis, we first approximate the signal-to-noise ratio (SNR) for a maximum-ratiocombining (MRC) receiver by a lognormal random variable in the presence of above channel impairments. We use two different techniques; the moment matching technique and the moment generating function (MGF)-based technique for lognormal approximation of the SNR and show that the two approximation techniques perform differently under different channel conditions. Based on the lognormal approximation of the SNR, we approximate the SINRs for MRC, zero forcing (ZF), and minimum mean squared error (MMSE) receivers by lognormal random variables and provide closed-form expressions for the outage probability and capacity of the system. The approximations are validated by comparing with Monte-Carlo simulations. The effects of shadowing on the performance of the system are quantified and it is illustrated that shadowing does not average out by increasing the number v