Abstract
An iterative decoding scheme for coded modulation systems over multipath fading channels is proposed to overcome the detrimental effects of intersymbol interference caused by the multipath channels. At the receiver, equalization and decoding process are jointly combined to apply the iterative decoding. The iterative decoding process is run over a factor graph, a cascade of the factor graph of a binary convolutional code and the graph a real-valued convolutional code (corresponding to the multipath channel). As in turbo decoding, the extrinsic soft information is extracted from the symbol detector and from the decoder at each decoding step and is used in the next decoding step. Performance analysis for multipath fast Rayleigh fading channel is provided. Simulation results are presented both for static multipath channel (ISI channel) and multipath fading channels. For the ISI channel, the intersymbol interference effect is completely overcome through the iterative decoding when SNR is high enough. For the multipath fading channel, although the ISI is not completely removed, the reduction of the intersymbol interference is remarkable and satisfactory for many practical applications.