AD3-GLaM: A cooperative distributed QoE-based approach for SVC video streaming over wireless mesh networks

We study the routing problem of scalable video coding video streaming over wireless mesh networks. In contrast to most of the conventional routing algorithms, our proposal focuses on optimizing users' satisfaction. The mean opinion score -an indicator of quality of experience (QoE) in video streaming- is utilized to assess the quality of routes in wireless mesh networks. The objective is to optimize the overall user experience in the network. Conventional routing approaches do not consider QoE and are not optimal with respect to user experience. Moreover, some centralized approaches are not scalable and require significant computational resources. The latter disadvantage can be overcome using distributed approaches. This paper presents a QoE-based cooperative distributed routing approach. Among distributed cooperative optimization schemes, AD3 is highlighted as one of the most efficient because of its fast convergence. The contributions of this paper are as follows: we encode the original problem into a factor graph and optimize the number of exchanged messages; we propose a partially distributed routing scheme based on OLSR and AD3; and we propose a distributed decoding algorithm in order to find a feasible solution. Our thorough simulation results confirm the advantages of the proposed scheme.