Efficient VCR-Aware VoD Services in P2P-Based Urban Vehicular Ad-Hoc Networks

Peer-to-peer (P2P) solutions have emerged as a successful approach for efficient multimedia resource distribution in vehicular ad-hoc networks (VANETs). However, it is a challenge to offer users interactive video-on-demand (VoD) services, mostly due to the difficulties in supporting VCR-like operations in such a highly dynamic mobile wireless network environment. In this paper, we propose a VCR-Aware VoD solution for P2P-based urban VANETs (VAVoD). VAVoD employs a grouping-based video segments storage solution to improve segment seeking success rate and load-balancing in Chord. Additionally, VAVoD makes use of a VCR-aware prefetching strategy to prefetch the segments predicted to be most likely requested by future VCR-like operations according to the current playing state. This strategy is based on a novel variable block mining and block jump association analysis algorithm for characterizing users' viewing behavior. VAVoD avoids the seriously long latency caused by relocating and fetching new video segments in mobile VANET environments when VCR occurs, and significantly ensures vehicle passengers' high levels of quality of experience (QoE). Simulation results show how VAVoD is a highly efficient interactive VoD solution in P2P-based VANETs in comparison to another state-of-the-art solution.