Multisample receivers for time-varying channels using particle filtering
Tadesse Ghirmai, Petar M. Djurić, Mónica F. Bugallo, Jayesh H. Kotecha · 2003
It has been reported that multisampling provides an implicit time diversity and that processing more than one sample per symbol results in achieving improved performance over processing a single sample. We explore this time diversity and develop a novel method for symbol detection from several samples per symbol. The method is based on a Bayesian formulation in which we use sequential Monte Carlo filtering, also known as particle filtering. Particle filtering has been applied to data detection problems with a single sample per symbol, and it has shown promising results. The proposed method is developed for communication systems characterized by time-varying channels. We demonstrate, via computer simulations, that significant performance improvement can be achieved by processing multisamples.