Quality-Driven Adaptive Video Streaming for Cognitive VANETs

Long Sun,Aiping Huang,Hangguan Shan,Min Xing,Lin Cai
DOI: https://doi.org/10.1109/VTCFall.2014.6966145
2014-01-01
Abstract:In cognitive vehicular ad hoc networks (CVANETs), channel conditions are highly dynamic due to both vehicle mobility and primary user activity. In this paper, to support high-quality video playback in such a challenging scenario, an adaptive video streaming algorithm built on scalable video coding (SVC) is proposed for reducing interruption ratio and improving visual quality. The proposed streaming algorithm is capable of deciding the proper number of video layers for vehicle users, by taking into account several important factors including vehicle position, velocity, the activity of primary users. Simulation results demonstrate the superiority of the proposed algorithm on playback interruption ratio and visual quality over the compared algorithm.
What problem does this paper attempt to address?