Iterative receivers with channel estimation for MIMO and multi-user OFDM systems

Sammanfattning: The traffic in wireless networks has been showing an exponential growth over the last decade. In order to meet the demand, and support a continuation of this growth, the scarce radio resources need to be efficiently used. The use of multiple antenna (MIMO) systems and iterative (turbo) receivers has significantly improved the realizable system performance. Still, many challenges exist in the development of efficient wireless receivers. In this thesis, which is written in the form of a collection of papers, we have investigated and developed iterative receivers with channel estimation for multi-user systems. Paper I investigates different iterative receiver algorithms for an uplink multi-user MIMO orthogonal frequency division multiplex (OFDM) system. For the given receiver structure, different combinations of multi-user detectors (MUD) and decision-directed channel estimators are compared. The comparison focuses on the complexity-performance tradeoff for the receiver configurations. The results show that low complexity algorithms, despite requiring more receiver iterations, tend to have the lowest overall complexity for a given bit error rate (BER). Paper II presents how the convergence behavior of an iterative receiver with channel estimation for an uplink multi-user MIMO-OFDM system can be modeled through multi-dimensional extrinsic information transfer (EXIT) charts. Although single-input single-output EXIT charts have been well studied, the multiple input-output case has generally been overlooked in the literature. Such charts are suitable for studying multi-user systems. The presented EXIT chart predictions show good correspondence with the true convergence behavior of the receiver. Paper III discusses the use of channel measurement data in system evaluations. We discuss and exemplify a number of issues related to the use of such data, and also provide potential solutions. The issues include problems with measurement noise and measurement antenna directivity. Paper IV investigates the performance gains that can be harvested through base station cooperation in an uplink multi-user MIMO-OFDM system. An iterative receiver with channel estimation is evaluated using unique mobile dual-link MIMO channel measurement data. The results point at the large performance gains which can be harvested through cooperative processing in a real environment. Paper V presents and evaluates a number of channel estimation algorithms for an OFDM interleave division multiple access (IDMA) system. We look at different pilot based algorithms as well as decision-directed algorithms being part of the iterative process of the receiver. As for paper I, the comparison focuses on the complexity-performance tradeoff, and the results show that low-complexity algorithms can be competitive alternatives when looking at the overall receiver complexity.