基于轨迹挖掘的公交车自组织网络路由机制  

A Routing Mechanism over Bus-Based VANETs by Mining Trajectories

作　　者：张扶桑[1,2] 金蓓弘[1] 汪兆洋 胡佳锋 张利锋[3] 

机构地区：[1]中国科学院软件研究所计算机科学国家重点实验室,北京100190 [2]中国科学院大学,北京100190 [3]鲁东大学信息与电气工程学院,山东烟台264025

出　　处：《计算机学报》2015年第3期648-662,共15页Chinese Journal of Computers

基　　金：国家自然科学基金(61472408;61372182)资助~~

摘　　要：数据传递是车载自组织网络(VANET)应用不可或缺的组成部分.在城市中,公交车网络具有覆盖面广、行驶线路固定等特点,VANET应用中的数据传递可以通过公交车自组织网络实现,但公交车网络能否在VANET数据传递中发挥其独特的作用取决于对公交车运行规律的挖掘和利用的程度.文中提出了一种公交车自组织网络的路由机制Vela,该机制依托从公交车历史轨迹中挖掘的公交车运行时空规律,建立了公交车数据传递的概率时空模型,并给出了基于路段上公交线路相遇的数据传递路径的选择策略.与已有相关研究相比,Vela路由机制一方面以路段为粒度刻画了公交车运行的时空规律,这使得所获得的公交车运行规律兼具准确性和稳定性,另一方面,基于公交车运行规律估算路段上数据传递的延迟和可靠性,能获得尽可能好的服务质量级别的数据传递路径.实验结果表明采用Vela路由机制能实现具有高到达率和低延迟的数据传递,而且Vela路由机制具有较强的可伸缩性.Data delivery is an indispensable component of applications over Vehicular Ad Hoc Networks（VANETs）.In urban scenarios,bus-based networks have distinctive features such as wide coverage and fixed lines so that they can be employed to achieve data delivery in VANET applications.However,whether a bus-based VANET can play its unique role in data delivery depends on how to effectively mine and utilize spatial-temporal rules implied in bus-based networks.This paper proposes a bus-based ad hoc routing mechanism Vela.By mining spatial-temporal rules from bus historical trajectories,Vela builds a probabilistic spatial-temporal model for data delivery and then gives the routing selection strategy based on patterns of bus line encountering on road segments.Compared with existing solutions,the routing mechanism Vela characterizes the spatial-temporal rules of buses in a moderate granularity of road segments,which makes the acquired spatial-temporal rules of buses both accurate and steady.On the other hand,Vela estimates data delivery delay and reliability based on the spatial-temporal rules,thus obtaining the routing paths with the best possible QoS（Quality of Service）levels for data delivery.Theexperimental results show that Vela performs much better in terms of delivery ratio and delay,and have the stronger scalability than other solutions.

关 键 词：车辆自组织网络 基于公交车的路由 轨迹挖掘 时间序列分析 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]