Iterative receiver design in Rayleigh fading using factor graph

In this paper we develop a novel approach for joint iterative channel estimation and LDPC decoding, based on the factor graph representation and sum-product algorithm. The fast flat fading channel is modeled as an AR process in order to characterize its temporal variation. By approximating the feedback message from the channel decoder as a Gaussian pdf, we show that Kalman smoothing can be employed for channel estimation. We further derive a new bit metric calculation method which consider both channel noise and channel estimation variance. Simulations confirm the effectiveness of our approach.