Computer Engineering Department, Computer and Information Systems College, Umm Al-Qura University, Makkah, Saudi Arabia.
PLoS One. 2022 Sep 16;17(9):e0273751. doi: 10.1371/journal.pone.0273751. eCollection 2022.
A vehicular network offers diverse beneficial services related to video streaming in different types of setups, including rural and urban. Some of the recent issues in vehicular communication include prospect of leveraging machine learning and blockchain for privacy and security enhancement, and resource allocation for video streaming coupled with integration of 6G networks for high data rate. Considering the extreme mobility and dynamic structure of vehicular networks and the high data rates of video streams, a unitary route may not support the required quality of a video stream. To achieve load balancing, connectivity among vehicles, path diversity, and low delay, the multipath transmission with a delay-tolerant network (DTN) concept based on a node disjoint algorithm is considered. In this proposed study, video frames are categorized in accordance with priority and forwarded via two graded paths. The first path carries the video reference frame, which is the most important frame for video decoding. The second path carries neighboring frames during video transmission. For the efficient selection of an optimal relay vehicle, a communication cost function is introduced into the existing DTN. This communication cost function is based on three key enhancement parameters: link stability rate, accessible bandwidth estimation, and transmission delay. The improvement in this study, is the integration of store-carry-forward strategy to the existing multipath data forwarding strategy. On the basis of the simulation outcomes, the proposed multipath video data communication in a vehicular DTN (MVDTN) scheme can enhance video data delivery in terms of packet loss ratio, end-to-end delay, structural similarity index measure, and peak signal-to-noise ratio. Considering the aforementioned metrics, our proposed schemes outperform the baseline schemes, namely, road-based multi-metrics forwarder selection evaluation for multipath video streaming and quality of service-aware multipath video streaming for an urban vehicular ad hoc network by using ant colony optimization.
车联网在不同类型的设置中提供了与视频流相关的各种有益服务，包括农村和城市。车对车通信中的一些最新问题包括利用机器学习和区块链来增强隐私和安全性，以及资源分配用于视频流以及与 6G 网络的集成以实现高数据速率。考虑到车联网的极端移动性和动态结构以及视频流的高数据速率，单一路径可能无法支持视频流所需的质量。为了实现负载平衡、车辆之间的连接性、路径多样性和低延迟，考虑了基于节点不相交算法的延迟容忍网络 (DTN) 概念的多路径传输。在这项拟议的研究中，视频帧根据优先级进行分类，并通过两条分级路径转发。第一条路径传输视频参考帧，这是视频解码最重要的帧。第二条路径在视频传输过程中传输相邻帧。为了有效地选择最佳中继车辆，在现有 DTN 中引入了通信成本函数。该通信成本函数基于三个关键增强参数：链路稳定性率、可访问带宽估计和传输延迟。这项研究的改进是将存储转发策略集成到现有的多路径数据转发策略中。基于仿真结果，在车对车 DTN 中提出的多路径视频数据通信 (MVDTN) 方案可以提高视频数据的传输效率在分组丢失率、端到端延迟、结构相似性指数测量和峰值信噪比方面。考虑到上述指标，与基于蚁群优化的城市车对车自组网的多路径视频流的基于道路的多指标转发器选择评估和服务质量感知的多路径视频流的基线方案相比，我们的方案表现更好。
