面向VANET可靠广播的协同背驮算法研究

发布时间：2018-01-10 21:15

  本文关键词：面向VANET可靠广播的协同背驮算法研究　出处：《计算机应用研究》2017年09期 　论文类型：期刊论文

  更多相关文章： 协同背驮 可靠广播 车辆信息孤岛 车载自组织网络

【摘要】：道路信号干扰造成的非可靠链路使得车载自组织网络(vehicular Ad hoc network,VANET)中的广播消息频繁发生数据丢包,由于广播消息中包含与车辆安全相关的信息,任何数据丢包都可能对车辆的安全驾驶造成严重影响。为增强车辆之间信息交互的可靠性,提出了一种贪心背驮算法(greedy piggyback algorithm,GPA)。允许所有车辆在发送自身丢包请求的同时,从接收缓存中选取若干消息进行背驮式附带广播,对其他车辆的数据丢包进行合作恢复。针对由非对称信道引起的车辆信息孤岛问题,在贪心背驮算法的基础上提出了一种基于策略学习的交互背驮算法(mutual learning and piggybacking algorithm,MLPA),通过车辆之间背驮策略的交互、学习提高VANET中消息广播的可靠性。仿真结果表明,GPA和MLPA不仅将车辆广播中数据接收的可靠度分别提高到0.85和0.95,还能获得较低的网络负载和较短的传输延迟。该算法可被推广至其他对实时性和可靠性要求较高的无线通信网络中。
[Abstract]:The unreliable link caused by road signal interference causes frequent packet loss in broadcast messages in vehicle Ad hoc Network (VANET). Because the broadcast message contains information related to vehicle safety, any data packet loss may have a serious impact on the safe driving of vehicles. In order to enhance the reliability of information interaction between vehicles. A greedy piggyback algorithm gregreedy piggyback algorithm is proposed, which allows all vehicles to send their own packet loss requests at the same time. A number of messages are selected from the receiving cache for piggyback incidental broadcasting to recover the data lost from other vehicles. The problem of isolated island of vehicle information caused by asymmetric channel is addressed. Based on the greedy piggyback algorithm, an interactive piggyback algorithm based on policy learning is proposed. Mutual learning and piggybacking algorithm. The reliability of message broadcasting in VANET can be improved by the interaction of carrier strategy between vehicles. The simulation results show that this method can improve the reliability of message broadcasting in VANET. GPA and MLPA not only increase the reliability of data receiving in vehicle broadcast to 0.85 and 0.95 respectively. The algorithm can be extended to other wireless communication networks with high real-time and reliability requirements.
【作者单位】： 南京林业大学汽车与交通工程学院;南京林业大学机械电子工程学院;
【基金】：江苏省高校自然科学基金资助项目(15KJB580009) 江苏省自然科学基金资助项目(BK20130977) 国家自然科学基金面上项目(11272159)
【分类号】：TN929.5;U495
【正文快照】： 0引言车载自组织网络(VANET)是由一定范围内的车辆通过自组织无线通信构成的分布式移动网络[1]。VANET中的各个车辆通过周期性广播CAM(cooperative awareness message),实现邻近车辆之间运行状态和道路信息的交互共享。如图1所示,由于车体对信号的遮挡以及植被绿化的多径衰落，

本文编号：1406823