Research Article
Software-Defined Live Video Delivery Networks with Device-to-Device Communications
@INPROCEEDINGS{10.4108/eai.13-7-2017.2270269, author={Dianjie Lu and Ren Han and Jie Tian and Weizhi Xu and Xiangwei Zheng and Guijuan Zhang and Hui Yu}, title={Software-Defined Live Video Delivery Networks with Device-to-Device Communications}, proceedings={Workshop on QoE-Aware Resource Allocation for Multimedia Communications}, publisher={EAI}, proceedings_a={QOE-RAMC}, year={2017}, month={12}, keywords={video delivery networks software-defined networking device-to-device communication multi-commodity flow}, doi={10.4108/eai.13-7-2017.2270269} }- Dianjie Lu
Ren Han
Jie Tian
Weizhi Xu
Xiangwei Zheng
Guijuan Zhang
Hui Yu
Year: 2017
Software-Defined Live Video Delivery Networks with Device-to-Device Communications
QOE-RAMC
EAI
DOI: 10.4108/eai.13-7-2017.2270269
Abstract
Video Delivery Networks (VDNs) have been widely applied to the video content delivery of mobile Internet recently. However, the traditional VDNs are all infrastructure pre-deployed which makes them inflexible to adapt to the diversity and mass of live video delivery in mobile Internet. Thus, how to meet mobile users’ live video demands while minimizing the delivery cost is still a challenge task. To solve this problem, we propose a Software-defined live Video Delivery Network (S-VDN) framework with Device-to-Device (D2D) communication for mobile Internet. With a global view of the network, S-VDN is capable of managing the mobile users’ requests and the network topology conveniently as well as scheduling the radio resources. In addition, S-VDN enables content delivery among closely located mobile users without going through base stations by employing the D2D technique. Furthermore, we formulate this optimization problem as a Multi-commodity Flow (MCF) problem and design two heuristic algorithms to obtain the optimal solution. Simulation experiments show that our proposed algorithms are efficient in terms of both performance and rental cost.
