An Adaptive Non-preemptive Scheduling Framework for Delay Bounded Traffic in Cellular Networks

Provisioning multimedia streaming services to mobile users in next generation wireless networks is considered critical to the successful deployment of such networks. Streaming traffic is characterized by the need of relatively high data transmission rates, and the need to limit the wireless network delays during transmission. Such factors contribute to the importance of the design and use of scheduling mechanisms that work at the streaming connection level to manage network resources. In this paper, we consider the problem of designing schedulers that aim at maximizing the achieved throughput subject to constraints on the maximum acceptable delay that can be tolerated by each traffic stream. We propose an adaptive scheduling framework for the non-preemptive delivery of traffic streams in cellular networks where a fixed number of channels are allocated to streaming services. The obtained simulation results indicate the competitiveness of the proposed design when used online to control traffic, and the usefulness of the underlying algorithms when used offline to analyze traffic traces