A Novel Multi-Step Opportunistic Scheduling Scheme for Streaming Applications over OFDMA Wireless Systems

摘要

In wireless networks, providing a high fairness allocation isbecoming a very challenging task. This paper has two objectives:firstly, we proposed a new optimization problem called SatisfactionProbability which consists in maximizing the number of satisfiedusers and secondly, we proposed a new scheduling scheme calledQoE-aware Multi-step Opportunistic (QEMO) scheduler. QEMO is builtin an extended cross-layer design. It takes in consideration notonly the transmission conditions but also the QoS experienced by theusers. Based on an opportunistic weighted scheme, our schedulertakes a maximum advantage of multiuser diversity optimizing fairnessand system capacity. This ensures an efficient support of multimediaservices in OFDMA wireless networks. Performance evaluation showsthat this well-balanced resource allocation scheme outperforms otherwell known opportunistic scheduler (MaxSNR, PF and RR) anddemonstrate that choosing between high system capacity and highfairness is not required.