In modern wireless networks base stations have been equipped with multiple antennas. When a mobile client also has multiple antennas, the additional antennas can be used to add extra data to a transmission signal, effectively increasing data rate or reliability through spatial multiplexing. Since the wireless channel generally adds unwanted noise to a signal sent over it, and this noise can vary in time, we effectively have a channel with time-varying quality. This channel quality has an effect on how fast data can be transmitted and in extension also on the operation of the transmit buffer behind it. It is the study of the dynamics of this buffer this project is interested in. The so-called multi-user MIMO (multiple-input, multiple-output) technology standardized in IEEE 802.11ac (a body for standardizing wireless communication) adds a dimension of complexity to the problem. Multiple users require different types of service, and this asymmetry creates some interesting buffer behaviour. The aim is to derive key performance measures such as delay, buffer content, or maximal achievable throughput, that are of interest to dimension the base station correctly. On top of that, the research project will study different transmission strategies in such a multi-user environment. The outcome of the project will be new stochastic models and queueing-theoretic analysis techniques, with practical applications in the performance evaluation of transmit buffers in a MUMIMO environment.