Details

Autor(en) / Beteiligte

• Anand, Arjun
• de Veciana, Gustavo

Titel

Invited paper: Context-aware schedulers: Realizing quality of service/experience trade-offs for heterogeneous traffic mixes

Ist Teil von

• 2016 14th International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks (WiOpt), 2016, p.1-8

Ort / Verlag

IEEE

Erscheinungsjahr

2016

Link zum Volltext

Quelle

IEEE Electronic Library (IEL)

Beschreibungen/Notizen

• Modern broadband wireless networks support application mixes, with different, possibly complex, application/user Quality of Service/Experience (QoS/QoE) metrics. The central problem underlying resource allocation for such systems is realizing QoS/QoE trade-offs given the dynamic loads and capacity variability they would typically see. The paper explores a framework for context-aware scheduling based on: (1) context-aware flow classification and management, and (2), complementary base station scheduler. Motivated by typical flow-size distributions for current traffic and characteristics of the associated delay optimal (Gittins index-based) schedulers we propose a novel flow and channel-aware scheduler which meets our design objectives. Using a combination of analysis and simulation we explore the achieved QoS/QoE trade-offs across a dynamic mix of traffic, in particular: 1) mobile web browsing and small file delays; 2) stored streaming video quality vs re-buffering; 3) throughput of larger file downloads. They suggest improved QoS/QoE tradeoffs vs traditional proportionally fair schedulers and are robust to the network load.