A novel analytical framework for mobility modeling in all-ip wireless systems

In all-IP wireless networks, groups of cells are served by IP access gateways. When users move between regions served by different gateways, IP and application layer mobility signaling is triggered. The key performance indicator to assess a given mobility protocol is the mean number of gateway handoffs. In this paper, we propose a new fundamental approach, which majorly extends past work in cellular communications, to evaluate the number of access gateway handoffs based on their size and the mobility pattern between cells during the session lifetime. Using transient Markov chain and complex analytic techniques, we obtain the mean number of handoffs for a session, be it entirely or only partially served by the network under consideration, for directed and random mobility. The main set of results indicates that the number of handoffs in the network is a non-linearly increasing function of the session duration and the number of access gateways, for both mobility patterns. To demonstrate the practical relevance of the approach we show results for MobileIP airlink load as a function of handoff rate.