Error propagation in decision feedback receivers

The problem of assessing error propagation in decision-directed feedback receivers is considered. It is shown that useful bounds on error statistics can be obtained by modelling the error mechanism using stochastic state-transition diagrams and then employing state reduction techniques to enable the complexity of the problem to be contained. This provides a more widely applicable approach to obtaining bounds on performance measures for such receivers than was previously available. The techniques are illustrated for representative error burst statistics of interest in practical applications.