Modelling time-limited capacity of a wireless channel as a Markov reward process | IEEE Conference Publication | IEEE Xplore