Finite-state Markov channel (FSMC) models are useful for analysing radio ch
annels with nonindependent fading, which give rise to bursty channel errors
. In the paper the FSMC model is constructed by partitioning the dynamic ra
nge of the fade amplitude into finite number of intervals and representing
each interval as a 'state' in the model. Transitions between these channel
states thus characterise the physical fading process. Based on the Nakagami
-m distribution, a new analytical formulation for the state transition prob
ability is presented and validated for channels with arbitrary degrees of t
ime-interleaving. The resultant FSMC model is shown to be applicable to a w
ide range of practical fading channels, including diversity-combined channe
ls and multicarrier channels.