Vehicular Network is a special self-organizing network and requires the communication pattern of"V2V" to share information among vehicles and meet the content request of various nodes. However, some vehicles prefer to receive content from others rather than send their own resources for the purpose to save their bandwidth. To solve this selfish behavior, credit-based incentive mechanism is often used but most of them only stimulate vehicles to"try but not try their best". This paper firstly proposes an innovative cooperative information sharing model in vehicular networks and focuses on the optimization mechanism. The server collects feedbacks from vehicles and use Dempster-Shafer theory to tell the truth from the false, thus finding more reliable cooperative vehicles. Simulation results show that the optimization mechanism could avoid malicious fake feedback information and perform well in raising stability and throughput of the vehicular networks.%车联网(Vehicular Networks)是一种车载自组织网络,需要节点之间的信息互享来保障网络内信息的传播,从而满足节点的内容需求.但是,很多节点为节省带宽等,往往只接收其他节点发送的数据,而不去共享自身数据.为了解决这种自私性带来的网络性能降低等问题,常常采用基于信用度的激励方式.然而,现有的信用度激励模型只解决了节点"发",而没有解决"尽力发"的问题.所以,在已有研究基础上,提出了一种新的多节点合作共享信息传输模型,并阐述了模型关键部分——合作节点选择优化方案.中央服务器通过收集节点的反馈信息,采用D-S证据理论(Dempster/Shafer Theory)区分虚假信息,实时评估节点的可靠性,从而提升信息传输的效率.此外,还提出了一种发送方和接收方双向确认的辅助机制,旨在源头减少虚假反馈的产生.仿真结果表明,该机制可以更加精准地选择出可靠合作节点,提升网络内信息传输的稳定性和吞吐量.
展开▼
