Power Allocation and Multiuser Scheduling for Delay-Limited Communications Exploiting Only Channel Statistics

This paper considers a delay-constrained wireless communication system in which delay constraint is represented as an outage rate probability constraint within a transmission of K-block Rayleigh flat-fading channels. Given a multiuser time-division (TD) multiplexing system with users´ delay constraints, our aim is to minimize the overall transmit power by optimizing the power allocation and the time sharing (i.e., the number of assigned blocks) among the users. The central transmitter (or scheduler) is assumed to have only the channel statistics (CST) while the mobile receivers (or users) have perfect instantaneous channel state information. This problem is of great importance for communication systems where the data rates are preset and the transmission cost is to be minimized. Our contributions are twofold. Firstly, we derive an analytic equation for the minimum power to support a given rate with an outage probability constraint if the number of assigned blocks for a user is known a priori. On the other hand, by constraining the upper bounds of the outage rate requirements, we show that the optimal time-sharing parameters can be found using convex optimization after relaxation. Our proposed solution hence first finds the time-sharing parameters by proper mapping from the relaxed optimum, and then uses the minimum power equations to obtain the corresponding optimal power distribution.