ABSTRACT
This chapter shows the importance of transient analysis in a time-varying system for which the steady state analysis may not provide very accurate performance measures. It presents an analytical model to investigate the transient quality of service performances in a multiservice code division multiple access (CDMA) mobile network. An application of the analytical model is presented to adaptively optimize the radio resource management parameters so that the system revenue can be maximized. Adaptive uniformization methods were proposed to analyze the transient state behavior of a Markov chain. The chapter reviews three different approaches (i.e., ordinal differential equation, matrix exponential, and uniformization) for transient analysis of a continuous-time Markov chain. It presents a framework for resource allocation and admission control in a CDMA cellular network based on the transient performance analysis. The framework can be extended to investigate both packet-level and call-level performance measures and their relationships in the transient state.
