Efficient detection and adaptive transmission for MIMO-OFDM systems

Orthogonal frequency division multiplexing (OFDM) is applied in many wireless communication standards for broadband data transmission, as it transforms the dispersive channel into a set of parallel non-interfering channels. Multiple input multiple output (MIMO) systems have become another key ingredient of modern communication standards. Multiple antennas allow for an improved link reliability due to spatial diversity and the simultaneous transmission of parallel data streams that can be separated at the receiver based on their different propagation conditions. The complexity of the optimal MIMO detector increases exponentially with the data rate. Therefore, a plethora of suboptimal detection algorithms has been proposed in the literature. This thesis provides a thorough overview of existing transmit strategies and detection algorithms and compares their performance in conjunction with channel coding. In this context, the importance of feedback from the channel decoder to the detector is investigated. In order to obtain general results that do not depend on a specific channel code, the main focus is on the achieved mutual information. If channel knowledge is available at the transmitter, it can adapt to the channel in order to improve the performance, in particular in combination with suboptimal receivers. For ideal channel state information, the MIMO channel can be diagonalized by adaptive filtering at the transmitter and receiver, while a low-rate feedback link only allows to adapt certain parameters like the data rate or the transmit power per antenna. Adaptive precoding also enables the simultaneous transmission of data to multiple users. This thesis provides a comprehensive survey of the advantages provided by channel state information at the transmitter.