Semi-Markov Modeling for Body Area Networks

Signal variation in Body Area Networks (BANs) is dominated by slow, macroscopic fading. The quantized (state) channel has long, and variable state-holding times. We motivate the use of a semi-Markov approach to model the quantized channel gain and show that for everyday activities and on-body (real-world) measurements this approach leads to tractable limits for state holding times and transition probability. We show that while activity does influence the parameters of the semi-Markov model, the impact is not strong.