A unified framework for adaptively scheduling hybrid voice/data traffic in 3G cellular CDMA downlinks

A unified framework for scheduling hybrid voice and data traffic is proposed for CDMA downlinks. We address the consistency of the framework as well as the distinctions of voice and data scheduling processes by discussing the common policy and individual requirements of both classes. An adaptive priority profile is designed in the scheduling algorithm based on queuing delay, required transmission power, and available transmission rate, which borrows the idea of composite metric from wired systems. With this design, the proposed algorithm accomplishes systems performance enhancement as a whole while retaining separate performance features without degradation. The uniformity of the proposed framework not only simplifies the implementation of the scheduling algorithms at base stations, but also is verified to be robust and resistant to various offered traffic load and variable service structure (voice/data proportion). Numerical conclusions show a system capacity gain of 4%-39% and traffic throughput improvement of 5%-26% most of the time over no scheduling systems. Compared with systems where no scheduling, or only data scheduling is employed, average outage probability is reduced by 50% and 94% respectively. The proposed framework is also tested to efficiently consume system resources by the approximate 100% power utilization capability, while no scheduling scenario exposes a 90% utmost.