VANETs中基于时空分析的抗合谋Sybil攻击检测方法

Spatio-Temporal Analysis Based Resist Conspiracy Sybil Attack Detection in VANETs

在车载自组网中,攻击者通过伪造、偷窃以及与其它合法车辆合谋等方式获得多个网络身份,并利用这些身份发布虚假交通信息来伪造交通场景,从而造成交通拥堵,甚至引发更严重的交通事故.现有的多数Sybil攻击检测方案主要用于检测伪造身份或偷窃身份的Sybil攻击,很少有检测方案针对合谋Sybil攻击进行研究.另外,检测Sybil攻击需要确保车辆实体在网络通信时仅绑定一个网络身份以防止攻击者扮演多个身份来欺骗其他车辆,这种做法会造成车辆隐私的泄露.为了平衡解决Sybil攻击检测和隐私保护这两个相互矛盾的问题,该文提出了一种基于时空分析的抗合谋Sybil攻击检测方法STARCS(Spatio-Temporal Analysis based Resist Conspiracy Sybil Attack).该方法利用匿名RSU发布的时间戳标识作为车辆在车载自组网中的身份,通过权威机构TA(Trust Authority)设置的请求信息表记录前一次经过的RSU和时间戳,从而抵御伪造Sybil身份和合谋Sybil身份的产生.文中每个警报事件只允许每个车辆发布一次带有时间戳标识的警报消息,根据警报信息中包含的标识是否被多个车辆同时使用来检测攻击者与合谋车辆同时使用同一身份的合谋Sybil攻击,并依据标识中嵌入的RSU位置关系标签检测攻击者与远距离车辆合谋的Sybil攻击.由于从时间上身份是否被滥用和空间上身份是否出现不合理位移的情况来检测合谋Sybil攻击,即是从时空关系上抵制和检测合谋Sybil攻击.理论分析和仿真实验表明,该方法不仅能够抵御或检测多种Sybil攻击,而且具有较少的时间开销和通信开销,并通过动态匿名机制保护车辆的身份和位置隐私.

In Vehicular Ad Hoc Networks(VANETs),Attacker can obtain multiple network identities by forging,stealing and conspiring with other legal vehicles,then they use these Sybil identities to transmit fake traffic information to create an illusion traffic congestion,even would cause more serious traffic accidents.The most current Sybil attack detection scheme primarily detects forged identity and stolen identity Sybil attack,there is rare research to detect conspiracy Sybil attack.In addition,Sybil attack detection need to bind each vehicle to a single identity in order to keep adversaries from playing multiple roles or spoofing with other legitimate identities,this will lead to privacy reveal issues.In order to balanced solve the two conflicting problems between Sybil attack detection and privacy protection,we propose a resist conspiracy Sybil attack detection method based on Spatio-Temporal Analysis based Resist Conspiracy Sybil Attack(STARCS).The method uses anonymous road side unit(RSU)released timestamp token as identity in VANETs,and there set a table to record RSU that last passed and timestamp by Trust Authority(TA),the table is called request information table to record the RSU and timestamp which vehicle last passed,to defend the generation of stolen and conspired Sybil identities.In this paper,each alarm event only allows each vehicle to release an alert message with timestamp token to detect Sybil attack,and we check whether the timestamp token contained in multiple alert messages and used by a plurality of vehicles at the same time to detect conspiracy Sybil attack which attacker and conspired vehicles using the same identity at the same time,and based on the adjacent relation tag of RSUs which embedded in the timestamp token to detect conspiracy Sybil attack which attacker conspired with the long-distance vehicles launch Sybil attack.Due to we detect conspiracy Sybil attack from the perspective of time to analyze whether the identity is abused or not and from the perspective of spatial to analyze whether the identity is appearing unreasonable displacement or not,that is,from the spatio-temporal analysis to resist and detect conspiracy Sybil attack.Theoretical analysis and simulation results show this method not only can defend or detect various types of Sybil attack,but also has less computation overhead and communication overhead,and it can protect the privacy of vehicle’s identity and position by dynamic anonymity mechanism.