An improved Markov model for the urban SOTM channel

In the time domain, the COTM channel is best modeled as a Markov generative process. The Baum-Welch parameter estimation algorithm is applied to empirical data collected in Boston, MA. The parameter estimation algorithm provides guidance for the development of a Markovian generative model to represent the ULMSC fade function in a larger time-based simulation. Three different continuous output transmission Markov models are presented and results are compared with discrete output Markov models. The continuous output models produce more accurate state durations, provide signal strength as a function of time and generally advance previous modeling efforts. The improved generative channel models also support the development of a larger time-based network simulation.