基于弱分类器集成的车联网虚假交通信息检测

doi:10.11959/j.issn.1000-436x.2016156

摘要/Abstract

摘要：

车联网中车辆以自组织的方式相互报告交通信息，开放的网络环境需要甄别消息，然而，要快速移动的车辆在短时间内检测出大量的交通警报信息是非常困难的。针对这一问题，提出一种基于弱分类器集成的虚假交通信息检测方法。首先，扩充交通警报信息的有效特征，并设计分割规则，将信息的特征集划分为多个特征子集；然后，根据子集特征的不同特性，使用对应的弱分类器分别进行处理。仿真实验和性能分析表明，选用弱分类器集成方法检测车联网中的虚假交通信息减少了检测时间，且由于综合特征的应用，检测率优于仅使用部分特征的检测结果。

关键词: 车联网, 虚假信息检测, 弱分类器集成, BP神经网络

Abstract:

Vehicles report traffic information mutually by self-organized manner in vehicular ad hoc networks (VANET),and the message need to be identified in the open network environment.However,it is very difficult for fast moving ve-hicles to detect a lot of traffic alert information in a short time.To solve this problem,a false traffic message detection method was presented based on weak classifiers integration.Firstly,the effective features of traffic alert information was extended and segmentation rules were designed to divide the information feature set into multiple feature subsets,then the corresponding weak classifiers were used to process feature subsets respectively according to the different character-istics of the subsets' features.Simulation experiments and performance analysis show that the selected weak classifiers integration method reduces the detection time,and because of the application of combined features,the detection rate is better than the test of using only some of the characteristics.

Key words: VANET, false information detection, weak classifiers integration, BP neural network

刘湘雯,石亚丽,冯霞. 基于弱分类器集成的车联网虚假交通信息检测[J]. 通信学报, 2016, 37(8): 58-66.

Xiang-wen LIU,Ya-li SHI,ENGXia F. False traffic information detection based on weak classifiers integration in vehicular ad hoc networks[J]. Journal on Communications, 2016, 37(8): 58-66.

图/表 12

图1

表1

表2

表3

图2

图3

图4

图5

图6

表4

图7

图8

参考文献 17

[1]	BAIOCCHI A , CUOMO F , FELICE D M , et al. Vehicular ad-hoc networks sampling protocols for traffic monitoring and incident detec-tion in intelligent transportation systems[J]. Transportation Research Part C:Emerging Technologies, 2015,56: 177-194.
[2]	常促宇，向勇，史美林 . 车载自组网的现状与发展[J]. 通信学报， 2007，28(11):116-126. CHANG C Y , XIANG Y , SHI M L . Present situation and development of vehicular ad hoc networks[J]. Journal on Communication, 2007,28(11):116-126.
[3]	ABUELELA M . A framework for incident detection and notification in vehicular ad-hoc networks[J]. Dissertations ＆ Theses-Gradworks, 2011.
[4]	GROVER J , LAXMI V , GAUR M S . Attack models and infrastructure supported detection mechanisms for position forging attacks in ve-hicular ad hoc networks[J]. CSI Transactions on Ict, 2013,1: 261-279.
[5]	RICHARD G E , MARTINE B , SAMUEL P , et al. VANET security surveys[J]. Computer Communications, 2014,44: 1-13.
[6]	GROVER J , PRAJAPATI N K , LAXMI V , et al. Machine learning approach for multiple misbehavior detection in VANET[M]. Advances in Computing and Communications.Berlin Heidelberg:Springer, 2011: 644-653.
[7]	ZHU W T , ZHOU J , DENG R H , et al. Detecting node replication attacks in wireless sensor networks:a survey[J]. Journal of Network and Computer Applications, 2012,35(3): 1022-1034.
[8]	ABDELAZIZ K C , LAGRAA N , LAKAS A , et al. Trust model with de-layed verification for message relay in VANETs[C]// The 2014 Interna-tional Conference on Wireless Communications and Mobile Comput-ing(IWCMC'14). c2014:700-705.
[9]	DING Q , LI X , JIANG M , et al. Reputation-based trust model in ve-hicular ad hoc networks[C]// The 2010 International Conference on Wireless Communications and Signal Processing (WCSP'14),Suzhou. c2010:1-6.
[10]	ZHANG J , CHEN C , COHEN R , et al. Trust modeling for message relay control and local action decision making in VANETs[J]. Security ＆Communication Networks, 2013,6(6): 1-14.
[11]	SHAIKH R A , ALZAHRANI A S . Intrusion-aware trust model for vehicular ad hoc networks[J]. Security ＆ Communication Networks, 2014,7(11): 1652-1669.
[12]	OSTERMAIER B , DOTZER F , STRASSBERGER M , et al. Enhancing the security of local danger warnings in VANETs-a simulative analysis of voting schemes[C]// The 2nd International Conference on Availability,Reliability and Security. c2007:422-431.
[13]	李春彦，刘怡良，王良民 . IP 车载自组网中基于交通场景的入侵行为检测机制[J]. 山东大学学报， 2014，44(1):29-34. LI C Y , LIU Y L , WANG L M . Intrusion detection scheme based on traffic scenarios in vehicular ad-hoc networks[J]. Journal of Shandong University, 2014,44(1):29-34.
[14]	KIM T H J , STUDER A , DUBEY R , et al. VANET alert endorsement using multi-source filters[C]// The 7th ACM International Workshop on Vehicular Ad Hoc Networks. c2010:51-60.
[15]	ZHANG J Y , HUANG L X , XU M J , et al. An incremental BP neural network based spurious message filter for VANET[C]// The 2012 In-ternational Conference on Cyber-Enabled Distributed Computing and Knowledge Discovery(CyberC'12).Sanya. c2012:360-367.
[16]	LIM S , HUIE L . Hop-by-hop cooperative detection of selective for-warding attacks in energy harvesting wireless sensor networks[C]// The 2015 IEEE International Conference on Computing,Networking and Communications (ICNC'2015). c2015:315-319.
[17]	詹珂昕 . 高速公路VANET预警信息传播机制的研究[D]. 北京: 北京交通大学 2012. ZHAN K X . 高速公路VANET预警信息传播机制的研究[D]. 北京: 北京交通大学 2012.

方法	基于节点特征的方法		基于信息特征的方法
方法	信任机制	投票机制	分类器优化	多源特征
优点	检测简单，开销小	能测出有历史行为的恶意节点	信息特征越多，检测精度越高	信息源越多，检测精度越高
缺点	受恶意节点的信任值影响	受恶意节点比例影响	检测复杂度高	依赖于环境因素
综合分析	检测简单，适用于车联网环境；依赖于通信环境，对消息判断往往不准确		信息特征和信息源越多，检测率越高；检测复杂度较高、警报信息较多时，时延会损害警报信息的时效性

序号	特征名称
1	Sender (s)
2	Priority (p)
3	Occur time (t₀)
4	Occur site (l₀)
5	Trans time (t_s)
6	Trans site (l_s)
7	Direction (d_s)
8	Vehicle vel (v_s)
9	Vehicle acc (a_s)
10	Reputation (r_s)
11	Retransmit (n)

特征子集	子集名称	子集内容
特征子集1	事件紧急度	t₀,l₀,p
特征子集2	道路状况	t_s,l_s,v_s,a_s,d_s
特征子集3	信息可信度	r_s,n

仿真参数	参数值
仿真时间	1 000 s
仿真场景范围	1 000 m×1 000 m
道路数量	Random
车辆数目	100辆
车辆速度	10~30 m/s
移动录制时间步长	0.05 s