Analysis methodology of adaptive resource allocation schemes for handoff calls: combining equilibrium and transient analysis

An important QoS issue in wireless networks is how to control handoff drops. There have been several adaptive admission control schemes that dynamically adjust the admission threshold based on handoff occurrences, but little effort for analysis has been made for it because of its mathematical modelling difficulties - it can be modelled as a Markov chain, but its huge state size makes the analysis intractable. We assume a basic adaptive admission control model and propose a novel analysis method adopting equilibrium and transient analysis together. This analysis methodology decomposes a Markov chain into the two levels, called intra-domain and inter-domain, to reduce the state size dramatically. Inter-domain is used for the equilibrium state analysis and intra-domain for the transient state one. They are used for finding handoff call dropping probabilities. By such decomposition, we can obtain the exact analysis of the system with much less matrix size.