Classification of Security attack in Vehicular Adhoc network: A survey

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.comVolume 2, Issue 2, March – April 2013 ISSN 2278-6856do not depend on fixed infrastructures and their nodes(known as mobile nodes) and may form networks on thefly for a variety of environments. Owing to the nature ofthese kinds of mobile nodes, VANETs can be challengedwith frequent topology changes as well as physical threats(that create potential vulnerabilities for potentialattackers). A successful attack on VANETs can havecatastrophic results (such as the loss of lives) or may leadto financial losses (for payment services). Thereforesecuring VANETs is crucial to the design,implementation and operation of these networks [1].Nowadays, transportation systems play an important partin our daily activities. Today major concern is to providesafety of users and safe their lives in road accidents.Safety and non safety potential applications of VANETare to ensure the safety of human life on the road.Security is the main concern of these applications where awrong message (due to insecure environment) maydirectly affect the human lives.Safety and non safety messages are forwarded betweenthe Vehicle to Vehicle (V2V) and Vehicle toInfrastructure (V2I) on this communication medium.Maxim Raya et.al [2] described the attacker model.Insider, malicious, active, and local is some example ofserious level attacks. Types of attacks can be differentdepending on the behaviour of attackers. Our focus pointis too safe the communication medium by controllingattackers. Active and passive safety applications [3] areused to provide safety to users. The role of active safety ishigh and it sends warning messages to other vehicles. Ifattackers change these messages then accidents become apart of the network and users’ life can be on risk.Each vehicle comprises of hardware unit for determiningcorrect location information using GPS. Roadside Units(RSUs) are placed across the road for infrastructurecommunication. The number of RSU to be used dependsupon the communication protocol.Several factors like the type of the road, daytime, weather,traffic density and even the driver himself affect themovements of vehicles on a road. Hence, the networktopology change frequently, and the routing protocol usedhas to adapt itself to these instantaneous changescontinuously.1.1. ScopeThe scope of this work is to study different types ofvehicular network security issues and theircountermeasures. The purpose of this paper is to find andstudy these vulnerabilities thoroughly and to present asolution to securing vehicular networks fromunauthorized and malicious access.2. Vehicular Ad Hoc Networks (VANET)The main idea behind VANET is providingcommunication among vehicles and also among vehiclesand a number of fixed equipment located on the road.The main intention with VANET’s is to enhancevehicles’ passengers’ safety and comfort by distributingtraffic, speed control, road and weather conditions amongnearby vehicles. Among its applications collisionwarnings, road sign alarms and automatic toll/parkingpayment can be mentioned.VANETs consist of a number of On-Board Units (OBU)which are located inside the vehicles and a number ofRoad-Side Units (RSU) which form the infrastructure ofthe network. OBU is a wireless gateway which can beconnected to other OBUs and RSUs. Each vehicleequipped with the OBU can become a part of the networkand will be able to send, relay and receive messagesthroughout the network. Due to its ad hoc nature, therewon’t be any centralized servers and vehicles are incharge of network management themselves [1].Besides safety applications VANET also provide comfortapplications to the road users.Communication in VANET is divided into two differentcategories:• Vehicle-to-Vehicle (V2V) Communication, were thenearby vehicles will communicate traffic and roadconditions to one another• Vehicle-to-Infrastructure (V2I) Communication, werevehicles communicate their gathered information to thenearest RSU in order to distribute the information fasterand more effectively.2.1 Security Requirements Of VANETIn order to have a secure and dependable vehicularnetwork, a number of security requirements must beconsidered. Some of these security requirements are thesame for all networks but some are valid and specific tovehicular networks only.With respect to the mentioned attacks and vulnerabilitiessecuring vehicular communications in all aspects is amust. Here we provide a list of some general securityrequirements that must be taken into consideration inorder to mitigate vulnerabilities and attacks againstVANETs.As far as security requirements are concerned theapplications of VANET are focused on safety messaging,cooperative driving, toll application etc. Therefore theintegrity, liability of message, liability of the user has tobe ensured and at the same time privacy has to be lookedupon. A secure VANET system should satisfy followingrequirements:2.1.1 AuthenticationDespite the lack of need for confidentiality, networknodes must be authenticated in order to be able to sendmessages through the network. Before reacting tomessages and events a vehicle must verify the legitimacyof the message and its sender, therefore there is a need forauthentication. Without authentication, illegitimate andmalicious users can inject false messages into the networkand confuse other vehicles by distributing falseVolume 2, Issue 2 March – April 2013 Page 372

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.comVolume 2, Issue 2, March – April 2013 ISSN 2278-6856information. With authentication, vehicles can simplydrop messages from unauthenticated users.2.1.2 AuthorizationAuthorization is on a higher level implemented by accesscontrol which itself is defined by network policies.Authorization defines the role of a node in the networkwhich includes the types of messages a node can read orwrite on the network, actions it is allowed to take andgenerally the protocols that it can execute.2.1.3 Data ConsistencyIn addition to authenticating the sender, the consistencyof messages with similar ones regarding time andlocation must also be considered, because false messagesfrom legitimate senders are not impossible. It is extremelyimportant for warning messages to meet the time andlocation constraints. A warning message must be shownto the driver before it is too late to react and also beforepassing the corresponding geographic location of thewarning.2.1.4 ConfidentialitySince security in vehicular networks is related to safety,all network users should normally have full access tonetwork data, i.e. traffic information, road conditions, etc.in order to make informed decisions. Since messages inVANETs don't contain any sensitive information and arenot confidential, there is no need for encryption andconfidentiality is not an important issue. Therefore,vehicular networks do not have to be protected againsteavesdropping. However, network data should be sentfrom authenticated sources and this can be done by sourceauthentication.2.1.5 IntegrityAll messages which are sent and received on the networkshould be protected against alteration attacks. A securevehicular network should provide protection againstmessage alteration. A message can be altered in severalways during its transit from source to destinations and allpossible attacks must be considered.When it comes to integrity there are three main threatsdirectly related to message contents. System threatsregarding integrity include (1) wrong or forged messages,(2) messages which are modified during transmission and(3) replayed messages.2.1.6 AvailabilitySince vehicular networks require real-time responses,they are vulnerable to DoS attacks. In order to remainoperational, protocols and services must be resilientagainst denial of service attacks. The communicationchannel must be available at all times, it must also bereliable otherwise attackers can launch DoS attacks. Suchattacks can disrupt the entire network which will lead tofailure in delivering network messages to other vehiclesin range. Therefore they must be securely designed and befault-tolerant in order to work under faulty conditions.There are a couple of security measures such as channelmonitoring which is a means to increase channelavailability. Besides, new technologies are always underdevelopment which will try to increase availability incommunication channels.Although availability can never be completelyguaranteed, yet there are still a couple of securitymeasures such as channel monitoring which is a means toincrease channel availability. Besides, channel switchingand using different communication technologies in caseof jamming attacks can be helpful.2.1.7 Non RepudiationWhen a node sends out a message, it shouldn’t be able tolater deny sending that message. In case of accidents orinvestigations, problem-causing drivers should be reliablyidentified, to correctly address the sequence and contentsof exchanged messages. This can be done by signingoutgoing messages with an anonymous key exclusivelyrelated to the sender and also a time- stamp associated tothe message preventing the user to claim that a particularmessage has been replayed.In case of using digital signatures, each message is signedwith its private but anonymous key. Therefore the vehicleowner cannot claim that he hasn't sent the message.2.1.8 PrivacyDriver privacy is an important issue in vehicularcommunications. Drivers don’t want their personal andprivate information to be accessible by others. Since thevehicle information such as location, speed, time andother car data are transmitted via wirelesscommunication, there should not be possible to infer thedriver's identity from this information. Among thisinformation, driver's location and tracing vehiclemovements are more sensitive and must be taken intoconsideration carefully .Regarding this issue we come to another requirementcalled “anonymity” which is discussed below.2.1.9 AnonymityAnonymity defines the requirement that network nodesmust not be able to infer if a node performed or willperform some specific action in future. In order to preventsuch inferences, there must be an equal probability ofdoing a specific action by all nodes, or have strongprobabilistic anonymity, with the probabilities beingequal for all nodes. By not requiring a vehicle toauthenticate with its exact identity to other vehicles towhich it sends information.2.1.10 Real-time ConstraintsIn order to have vehicular networks secured, specificmeasures and techniques have been suggested. Althoughthese techniques seem applicable and useful, there is oneimportant issue that must be taken into account, real-timeVolume 2, Issue 2 March – April 2013 Page 373

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.comVolume 2, Issue 2, March – April 2013 ISSN 2278-6856constraints. Out-dated messages, e.g. out-dated traffic orroad/weather conditions, are of no use in vehicularnetworks and must be eliminated in order to let the newlygenerated messages get to their destinations on time. Onthe other hand, solutions such as digital signatures andcertificates must be designed specifically for vehicularnetworks in order to fulfill real-time constraints of suchnetworks.3. Security challenges in VANETsDespite the various benefits offered by VANETs, theyintroduce several security challenges to the researchcommunity. These security challenges are mostlyconcerned with the trade-off between authentication andnonrepudiation versus privacy during communicationwithin VANET environments. Another security challengeis delay sensitivity because significant delays prohibit theuse of security protocols that have high overheads or relyon multiple stages of full-duplex communication betweennodes [4,5].The use of traditional authentication mechanisms toaddress some security threats (such as illusion attacks)without affecting privacy in VANETs poses a newchallenge for those VANET applications that need toauthenticate nodes in VANETs [6,7].The other challenge is the restriction introduced by thetransitory nature of interactions in a VN to usereputation-based schemes [8].Another VANET security challenge is the forwarding ofevent-related messages on very large ad hoc networks ofhighly mobile nodes in such a way that the informationcan be trusted by receiving nodes because by doing so weimprove traffic safety and mobility [9].All these challenges need to be addressed to provide asecure VANET infrastructure where nodes (vehicles) cancommunicate securely with each other and infrastructureunits.Table 1 Summary of security attacks in VANETs11 This survey paperidentify majorsecurity issues andchallenges associatedwith different routingprotocols.Third Type: TimingAttackForth Type: SocialAttackFifth Type:Monitoring AttackFirst Type: Attacks onavailability1.Black Hole Attack2.Malware3.BroadcastTampering4.Spamming5.Greedy Drivers6.Denial of ServiceSecond Type: Attackson Authentication/identification1.Masquerading2.Replay Attack3.Global PositioningSystem (GPS)Spoofing4.Tunneling5.Sybil AttackRef.ObjectiveNo.10 In this paper, wepropose five differentclasses of attacks andevery class isexpected to providebetter perspective forthe VANET security.The maincontribution of thispaper is the proposedsolution forclassification andidentification ofdifferent attacks inVANET.Types of AttackFirst Type: NetworkAttack1. Denial of service(DOS) Attack2.Distributed Denial ofservice (DDOS) Attack3.Sybil Attack4.Node ImpersonationAttackSecond Type:Application Attack(AP)12 This paper aims todescribe and analyzethe mostrepresentative6.Message Tampering7.ID DisclosureThird Type: Attackson confidentialityFourth Type: Attackson privacyFifth Type: Attacks onnon-repudiationSixth Type: Attackson data trustFirst Type : Attackson identification andauthenticationVolume 2, Issue 2 March – April 2013 Page 374

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.comVolume 2, Issue 2, March – April 2013 ISSN 2278-6856VANET securitydevelopments1.Impersonation.1.1False attributepossession.2. Sybil.Second Type: Attackson privacy1.Identity revealing.2. Location tracking.Third Type: Attackson non-repudiationFourth Type: Attackson confidentialityFifth Type: Attacks onavailability1.Network Denial ofService (DoS)1.1Routing anomalies.13 In this paper, wehave classifieddifferent securityattacks and logicallyorganized/represented in a more lucidmanner based on thelevel of effect of aparticular securityattack on intelligentvehicular traffic.1.Malicious nodes2.Illusion attackFifth Type: Location1.Forging positionsand Sybil attacks2.Position cheatingand false positiondisseminatingFirst Type: NetworkAttacks1.Denial of service(DOS) Attack2.Distributed Denial ofservice Attack3.Sybil Attack4.Node ImpersonationSecond Type:Application Attack2.Computation DoS.Third Type: TimingAttack1 This paper discussessome of the mainsecurity threats andattacks that can beexploited inVANETs and presentthe correspondingsecurity solutionsthat can beimplemented tothwart those attacks.Sixth Type: Attackson data trustFirst Type:Anonymity1.Malicious VehicleSecond Type: Keymanagement1.Brute force attack:2. Misbehaving andfaulty nodesThird Type: Privacy1.Malicious user2.Traffic analysisattacksFourth Type :Reputation14 Finally, we outlinesome of the VANETresearch challengesthat still need to beaddressed to enablethe ubiquitousdeployment and widespread adoption ofscalable, reliable,robust, and secureVANETarchitectures,protocols,technologies, andservices.Fourth Type: SocialAttackFifth Type:Monitoring AttackFirst Type: Threats toavailability1.Denial of ServiceAttack2.BroadcastTampering3.Malware4.Spamming5.Black Hole AttackSecond Type: Threatsto authenticity1.MasqueradingVolume 2, Issue 2 March – April 2013 Page 375

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.comVolume 2, Issue 2, March – April 2013 ISSN 2278-68562.Replay Attack3.Global PositioningSystem (GPS)Spoofing4.Tunneling5.Position Faking6.Message Tampering7.MessageSuppression/Fabrication/ Alteration8.Key and/orCertificate Replication9.Sybil AttackThird Type: Threatsto confidentiality4. PROPERTIES OF ATTACKERSAttacker create problem in the network by getting fullaccess of communication medium DSRC. Here we arediscussing some properties and capability of the attackerswhich has been mentioned in studies.InsiderThis type of attackers who is an authentic user of thenetwork and have detail knowledge of network. Insiderattacker might have access to insider knowledge and thisknowledge will be used for understanding the design andconfiguration of network. When they have all informationabout the configuration then it’s easy for them to launchattacks and create more problem as compare to outsiderattacker. It can create problem in the network bychanging the certificate keys. We can simply say thatinsider attacker is the right man doing the wrong job inthe network.OutsiderThe outsider attacker is considered as an authentic user ofthe network. It is a kind of intruder which aims to misusethe protocols of the network and the range of such attacksare limited. Outsider attacker also has a limited diversityfor launching different kind of attacks as compare toinsider attacker.Coverage areaCoverage area is the main property of attacker when theylaunch any kind of attacks. Attacker could cover the mainarea of road, and it depends on the nature of the attacks.Basic level attacker has controlled one DSRC channelsand covers the range of at most 1000 meters but theextended level attackers are more organized and covermore area using of hundred DSRC channels.Technical ExpertiseTechnical expertise of the attacker makes them strongerfor creating attacks in the network. It is difficult forattacker to mount attacks on cryptographic algorithms.Chance is low for attacker to compromise theinfrastructure network and data capture from restrictedarea of network. Attacker having ability to extracts theprogram code and secret keys of the computing platformof OBU and RSU by launching physical attacks.ResourcesBudget, manpower and tools are the three main keyresources and attackers depend on it to achieve theirgoals. Need budget to borrow technical expert and spendtime to understand the configuration of specific networkand then disturb network with launching of different kindof attacks. Attacker can use different kind of tools forlaunching attacks. These software tools can develop byown self or buy from the market. Many business partiesmake setup their business nears the road and provide nonsafety application services (Internet, entertainmentservices). One business party can be used their ownmaximum resources to create problems for other partiesand destroy their business with different kind of attacks.5. Conclusion and Future workIn this paper various attacks in VANET have beenclassified. It has been observed that the classificationhelps to deal with different types of attack on VANET.Due to recent improvements in connecting vehicles toexternal networks, proper security mechanisms must bedeveloped in parallel to reduce the risk of malicious andunauthorized behavior in the vehicular network domain.Users require safety on road in future vehicular networkand it could be possible by implementing VANETapplications.We expect that the proposed attacks classes may helpfulto identify attacks and understand attackers. It is difficultto control attackers but in future work we will developsuch system to identify attacks in network with respect tosome specific type of attack. Implementation could beeasy of this future human life saving network if wecontrol attackers and their attacks.In this work, we have analyzed the attacks that VANETscan be subjected to. Table provides a summary of thesecurity attacks we have identified that may be launchedon VANETs and the corresponding security solutions tomitigate those attacks. I hope that this survey will enableVANET designers and developers to build more secureand robust VANET architectures, protocols andapplications in the future.Volume 2, Issue 2 March – April 2013 Page 376

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.comVolume 2, Issue 2, March – April 2013 ISSN 2278-6856REFERENCES[1] J.T. Isaac S. Zeadally J.S. Ca´mara “Security attacksand solutions for vehicular ad hoc networks” IETCommunications 2009.[2] M. Raya, J. Pierre, Hubaux,”Securing vehicular adhoc Networks” Journal of Computer Security,vol.15,january 2007, pp: 39-68.[3] A.Weimerskirch,J.JHaas,Y.C.Hu,K.P.Laberteaux,”Data security invehicular communication networks”,chapter no.09,pp309-310.[4] BURMESTER M., MAGKOS E.,CHRISSIKOPOULOS V.: ‘Strengthening Privacyprotection in VANETs’. IEEE Int. Conf. on Wireless& Mobile Computing, Networking &Communication (WIMOB 2008), 2008, pp. 508–513[5] JAKUBIAK J., KOUCHERYAV Y.: ‘State of the artand research challenges for VANETs’. Fifth IEEEConsumer Communications and Networking Conf.(CCNC 2008), 2008, pp. 912–916[6] XI Y., SHA K., SHI W., SCHWIEBERT L., ZHANGT.: ‘Enforcing privacy using symmetric random keysetin vehicular networks’. Eighth Int. Symp. onAutonomous Decentralized Systems (ISADS 2007),2007, pp. 344–351[7] LO N., TSA H.: ‘Illusion attack on VANETapplications – a message plausibility problem’. IEEEGlobecom Workshops, 2007, pp. 1–8[8] PARNO B., PERRIG A.: ‘Challenges in securingvehicular networks’. Fourth Workshop on Hot Topicsin Networks (HotNets-IV), 2005[9] FLORIAN D., LARS F., PRZEMYSLAW M.:‘VARS: a vehicle ad hoc network reputation system’.Int. Conf. on a World of Wireless, Mobile andMultimedia Networks (WOWMOM 2005), 2005, pp.454–456[10] Irshad Ahmed Sumra,Iftikhar Ahmad, HalabiHasbullah “Classes of Attacks in VANET”[11] Anup Dhamgaye, Nekita Chavhan “Survey onsecurity challenges in VANET” IJCSN InternationalJournal of Computer Science and Network, Vol 2,Issue 1, 2013 88[12] José María de Fuentes, Ana Isabel González-Tablas,Arturo Ribagorda “Overview of security issues inVehicular Ad-hoc Networks”[13] Adil Mudasir Malla Ravi Kant Sahu “SecurityAttacks with an Effective Solution for DOS Attacksin VANET” International Journal of ComputerApplications (0975 – 8887) Volume 66– No.22,March 2013[14] Sherali Zeadally · Ray Hunt · Yuh-Shyan Chen ·Angela Irwin · Aamir Hassan “Vehicular ad hocnetworks (VANETS): status, results, and challenges”© Springer Science+Business Media, LLC 2010Volume 2, Issue 2 March – April 2013 Page 377