No abstract

No abstract

No abstract

Real-time video dissemination over Vehicular Ad hoc Networks (VANETs) is fundamental for many services, e.g., emergency video delivery, road-side video surveillance, and advertisement broadcasting. These applications deal with several challenges due to strict video quality level requirements and highly dynamic topologies. To handle these challenges, geographic receiver-based beacon-less approaches have been proposed as a suitable solution for forwarding video flows in VANETs. In general, the routing decisions are performed only based on network, link, and/or node characteristics, such as link quality and vehicle's location. However, in real situations, due to different requirements and hierarchical structures of multimedia applications, these existent routing decisions are not satisfactory to select the best relay nodes and build up reliable backbones to delivery video content with reduced delay and high Quality of Experience (QoE). This paper introduces the QOe-Driven and LInk-qualiTy rEceiverbased (QOALITE) protocol to allow live video dissemination with QoE assurance in Vehicle-to-Vehicle (V2V) scenarios. QOALITE considers video and QoE-awareness, coupled with location and link quality attributes for relay selection. Simulation results show the benefits of QOALITE when compared to existing work, while achieving multimedia transmission with QoE support and robustness in highway scenarios.

Vehicular Ad Hoc Networks (VANETs) promise a wide scope of real-time multimedia services, which need to deal with strict quality level requirements and highly dynamic network topology. To handle these challenges, multipath routing approaches have been applied to bring improvements in Quality of Experience (QoE) levels and support on-road real-time video delivery. Though, in real situations, there will likely be multiple streams simultaneously, which can cause more congestion periods and packet loss rates. This paper proposes the Multi-flow-driven VIdeo DElivery (MVIDE) mechanism to select best routes for live video sequences in VANETs. MVIDE can be integrated with routing protocols to define routes considering the characteristics of multiple paths, vehicle mobility, and application requirements. This cross-layer mechanism ranks the quality of candidate paths based on the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method and, subsequently, assigns paths to different substreams. MVIDE was added to the GPSR-MA protocol, being called GPSR-MA-MVIDE. Simulating results show that GPSR-MA-MVIDE outperfoms GPSR-MA in terms of multi-flow handling and received video quality.