Q-POINT: QoE-Driven Path Optimization Model for Multimedia Services

• Published in: Wired/Wireless Internet Communications pp. 134 - 147
• Main Authors: Dobrijevic, Ognjen, Kassler, Andreas J., Skorin-Kapov, Lea, Matijasevic, Maja
• Format: Book Chapter
• Language: English
• Published: Cham Springer International Publishing 2014
• Series: Lecture Notes in Computer Science
• Subjects: mixed integer linear program; Multimedia services; network-wide optimization; quality of experience; software-defined networking
• ISBN: 3319131737; 9783319131733
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-319-13174-0_11

When delivering multimedia services over Internet, different media types are impacted by resource limitations in a different way. While an interactive audio service calls for low-latency communication, video streams should be routed over network paths with sufficient capacity. However, in current networks flows towards the same destination follow the same path, which may lead to a suboptimal resource utilization that effectively penalizes end-users’ quality of experience (QoE). This paper proposes Q-POINT, a QoE-driven path optimization model to fairly maximize aggregated end-user QoE for competing clients’ service flows by calculating the best path for each flow, subject to resource constraints. We formulate the problem as a mixed integer linear program integrating QoE models for audio, video and data transfer. Such an approach can be leveraged within the software-defined networking paradigm, which provides a control plane to orchestrate path set-up. We evaluate our model and illustrate its benefits over shortest path selection.