Journal
IEEE TRANSACTIONS ON MULTIMEDIA
Volume 16, Issue 3, Pages 848-863Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TMM.2014.2300442
Keywords
Cross-layer; fairness; OFDMA; resource allocation; scalable video coding (SVC); wireless networks
Categories
Funding
- European Commission under the FP7 ICT (CONCERTO) project [288502]
Ask authors/readers for more resources
The design of optimized video delivery to multiple users over a wireless channel is a challenging task, especially when the objectives of maximizing the spectral efficiency and providing a fair video quality have to be jointly considered. In this paper we propose a novel cross-layer optimization framework for scalable video delivery over OFDMA wireless networks. It jointly addresses rate adaptation and resource allocation with the aim of maximizing the sum of the achievable rates while minimizing the distortion difference among multiple videos. After having discussed the feasibility of the optimization problem, we consider a vertical decomposition of it and propose the iterative local approximation (ILA) algorithm to derive the optimal solution. The ILA algorithm requires a limited information exchange between the application and the MAC layers, which independently run algorithms that handle parameters and constraints characteristic of a single layer. In order to reduce the overall complexity and the latency of the optimal algorithm, we also propose suboptimal strategies based on the first-step of the ILA algorithm and on the use of stochastic approximations at the MAC layer. Our numerical evaluations show the fast convergence of the ILA algorithm and the resulting small gap in terms of efficiency and video quality fairness between optimal and suboptimal strategies. Moreover, significant individual PSNR gains, up to 7 dB for high-complexity videos in the investigated scenario, are obtained with respect to other state-of-the-art frameworks with similar complexity.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available