Resource control for loss-sensitive traffic in CDMA networks

We investigate the problem of efficient resource control for loss-sensitive traffic in CDMA networks, using an economic modelling framework that takes into account the joint control of the transmission rate and the signal quality. Although the corresponding global optimization problem has a non-trivial structure, and we cannot in general guarantee that a solution can be found using the Lagrangian method, we have strong experimental evidence that this is possible for a wide range of user utilities. In this case the global optimum can be achieved in a decentralized manner, using shadow prices to influence the individual user resource requests. Based on this evidence, the main contribution of the paper is to discuss how existing rate control and outer-loop power control procedures can obtain a simple and attractive form that takes into account, through shadow prices, the level of demand and supply in order to achieve efficient resource utilization. Moreover, we describe and evaluate approximations of the proposed resource control model that can simplify its application, and we present extensions of the model for the case where the packet success ratio depends on the transmission rate in addition to the signal quality, and for network paths containing multiple wireless links.