基于信息中心5G车联网中社会感知的流媒体缓存与转发策略  

Social-aware caching and forwarding for video streaming in information-centric 5G VANET

作　　者：曹腾飞[1] 江翠丽 刘志强[1] 王晓英[1] CAO Tengfei;JIANG Cuili;LIU Zhiqiang;WANG Xiaoying(Department of Computer Technology and Applications,Qinghai University,Xining 810016,China)

机构地区：[1]青海大学计算机技术与应用系

出　　处：《电信科学》2019年第12期90-98,共9页Telecommunications Science

基　　金：青海省应用基础研究计划资助项目(No.2019-ZJ-7034);青海大学实验教育教学研究项目(No.SY20907);国家自然科学基金资助项目(No.61762074)~~

摘　　要：5G通信技术的到来为车载自组网(VANET)提供高质量的视频流服务奠定了基础。然而,现有的VANET仍然采用以主机通信而非内容分发的IP网络来分发视频数据,造成了网络功能与设计目标之间的不匹配,为提供高质量的视频服务带来了巨大挑战。基于上述问题,提出了一种以信息为中心的社会感知流媒体缓存和转发策略(SACF),其时间复杂度为O(m2)。首先,通过分析用户节点的行为构建社会感知的虚拟社区;然后,提出了一种视频快速转发机制来支持用户就近获取视频内容;最后,提出了一种基于社区的缓存策略来优化视频的缓存分布。仿真实验结果表明,与最新的解决方案相比,该算法的缓存命中率提高了18%,查找时延降低了33%,平均网络开销减少了14.6%。The arrival of 5G-enabled communication technology has laid the foundation for providing high-quality video stream services in vehicular Ad Hoc network(VANET). However, current VANET still distribute video content over IP networks which design for host communication instead of content distribution. The mismatch between network function and design goals brings huge challenges for providing high quality video service. Based on the above questions, a social-aware caching and forwarding was proposed to support efficient video streaming services and the time complexity of the algorithm was O(m^2). Firstly, a virtual community was constructed by analyzing the user’s behaviors. Then, a fast forwarding scheme was introduced to ensure that users could discover the requested videos nearby. Finally, a community-based caching policy was proposed to optimize the in-community caching distributions. Simulation results show how the proposed strategy achieves better performance than other state-of-art solutions in terms of caching hit ratio, searching delay and control overhead.

关 键 词：社会感知 缓存与转发 视频流 车联网 信息中心网络 

分 类 号：TP393.4[自动化与计算机技术—计算机应用技术]