A forward-backward Kalman filter-based STBC MIMO OFDM receiver
EURASIP Journal on Advances in Signal Processing
| Hi-index | 0.00 |
Accurate channel state information is important in communication systems. This is especially challenging in a wireless environment where the channel exhibits strong frequency and time selectivity. The literature is full with ingenious works devoted to the design and analysis of algorithms for channel estimation. In general, these works have approached the various algorithms distinctly obscuring commonalities that might exist among them. This dissertation presents two contributions related to the analysis and design of adaptive channel estimation algorithms. The first part of the dissertation performs an analysis of a large class of adaptive algorithms for channel estimation. Adaptive filters are, by design, time-variant, nonlinear, and stochastic systems. For this reason, it is common to study different adaptive schemes separately due to the differences that exist in their update equations. The dissertation presents a unified approach to the analysis of adaptive filters that employ general data or error nonlinearities. In addition to deriving earlier results in a unified manner, the approach presented also leads to new stability and performance results without imposing restrictions on the color or statistics of the input sequence. The second part of the dissertation presents an expectation-maximization (EM) based class of algorithms for joint channel and data recovery in OFDM (orthogonal frequency division multiplexing). The algorithms make use of the rich structure of the underlying communication problem—a structure induced by the data and channel constraints. These constraints include pilots, the cyclic prefix, the code, and the finite alphabet constraints on the data; sparsity, finite delay spread, and the statistical properties of the channel (time and frequency correlation). The algorithms become progressively more sophisticated as more data and channel constraints are incorporated, with each new version of the algorithm subsuming the previous version as a special case, culminating in an EM-based forward backward Kalman filter. The dissertation finally scales up the algorithm design to support OFDM transmission over multiple-input multiple-output (MIMO) systems.