FMO frame selection using markov model prediction in H.264 for slow fading wireless channels

In this paper we propose an adaptive FMO selection scheme using markov models to predict the future condition of the channel in order to further improve the performance of transmitting H.264/AVC video stream over wireless channels with FMO. Based on the packet errors of the previous frame a channel prediction is made using markov models in order to determine the number of slice groups that will be used to encode the next frame. Experiments show that for slow fading channels with short burst lengths the future channel conditions can be effectively modeled using markov models. Using the markov model we tried to modify the number of FMO slice groups that will be used to encode the current frame. We showed that for low-delay and low bandwidth video transmission with adaptive FMO slice group encoding we can improve the decoded video quality as compared to using FMO with fixed number of slice groups.