A Markov renewal based model for wireless networks (Q1598897)

In wireless networks the integration of the mobility and the teletraffic aspects is a crucial topic. The authors model the space as a finite set of cells, the mobiles arrive according to a non-homogeneous Poisson process. The succesive cells visited by a mobile form a discrete time Markov chain and the cell residence times have a general distribution which depends on the actual cell and both, the previous cell and the next cell to be entered, too. This allows a more path-oriented restriction of mobility, f.e. one-way highway traffic. The evolution of the teletraffic state of a mobile (customer) is represented by a Continuous Time Markov Chain (CTMC). The system is modelled as an infinite server model (no capacity constraints). Additionally each customer is attached an attribute (label, type,\dots) that changes with time by a CTMC, but independent of the actual location. The authors analyse the transient and limit number of customers and get independent Poisson random variables. They calculate covariances and show that the arrival processes (to a node) are sums of independent Poisson cluster processes. Specific specializations for wireless networks are illustrated and analyzed finally.