A Scalable and Secure Key Distribution

Scheme for Group Signature based

Authentication in VANET

Kiho Lim, Kastuv M. Tuladhar,

Xiwei Wang, Weihua Liu

Outline

Introduction to Vehicular Networks

Motivation

System Model

Secure Key Distribution Scheme

Evaluation & Analysis

Conclusion

Introduction to Vehicular

Networks

• Vehicular Ad hoc Networks (VANET) are a

special type of Mobile Ad hoc Networks

(MANETs) with fast-moving vehicular nodes,

dynamically changing topology.

• VANETs allow vehicles to share safety and traffic

information such as weather information,

broadcasting emergency condition, navigations

etc.

Vehicular Networking

Source: “Vehicular Networking: Standards, Protocols, Applications, and Deployment Plans”, Tutorials, 2015 IEEE CCNC

• Individual functions (like detecting seat occupancy)

are performed by individual embedded systems, called

Electronic Control Units (ECUs)

– essentially a ruggedized miniature computer,

typically based on a microcontroller, that can gather

data from attached sensors or trigger attached

actuators

Electronics in the vehicle

Extension to Vehicle-to-Vehicle

Communications

The next stage is to connect the in-vehicle network of

a car with other cars, as well as with its environment,

to open a whole set of new possibilities

Typical Connections

Specifically designed protocols between Roadside

Units (RSU) and vehicles; and vehicles to vehicles:

IEEE WAVE, IEEE 802.11p, ITS-G5, ARIB T109

Vehicular Networking Objectives

• Safety-relevant information can be exchanged that can

not be obtained using local sensors

– Enables the driver to “see through and ahead”other vehicles, buildings and environment

• Potential to improve driving comfort

• Potential to reduce congestion and thus environmental

impact

• And, sometime in the future, enable fully automated

rides

Cooperative Safety Systems

– Some Examples

Security Issues

• Propagation of falsified warning messages

can mislead towards an accident and

damage the life/property.

• Various Attacks in VANETs:

• bogus information attacks

• spoofing attacks (Identity, speed, position)

• privacy attacks

/17

Real World Attack #1: RollJam

unlock #n

unlock #n+1

unlock #n+2

……

Rolling Code for Keyless Entry

/17

Real World Attack #1: RollJam

unlock #n+1jam and capture #n+1

not opened!unlock #njam and capture #n

opened!replay #n

replay #n+1opened!

/17

Real World Attack #2: VW RKE HackIn USENIX Security 2016, Garcia et al. present that

only 4 encryption keys are universally used over

100M vehicles produced by VW group over the 20

years.

1

3

/17

Real World Attack #3: FCA Jeep Hack

• After the first release in 7/21/2015, it drew significant media

attention.

• FCA Jeep Cherokee “remotely” controlled by Charlie Miller

and Chris Valasek.

• On 7/24/2015, FCA issued a recall to 1.4M vehicles.

FCA Jeep Cherokee

FCA SDN

TCU

BCMPCM

CAN bus

IVIHackers

/17

Real World Attack #4: Remote Tesla Model S Hack

Motivation

• It is difficult to store/manage all keys in a vehicle.

• Centralized trusted authority has high burden of

generating and managing the group public/private keys.

• Another challenge in VANETs is delivering group private

keys securely from the key generator to vehicular nodes.

• A group is confined to the coverage of a road side unit

(RSU).

• Thus, our goal is to mitigate frequent key updates

requirement and to make the key management process

more efficient and scalable.

Related Works• Chaum et al. [4] introduced group signatures for anonymous

authentication, which employs several group keys

corresponding to one group public key.

• Sun et al. [14] proposed a pseudonymous authentication for

vehicular communication to provide anonymity and

traceability.

• A distributed key management framework [5] distributes the

group key with the help of RSUs.

• However, frequent key establishment has not been

addressed.

• Also, delivering the group keys in a secure manner is crucial.

System Model

Overview of the System Model

System Model• Trusted Authority (TA): Vehicles are registered by the

trusted authority and provided the certificates. TA and RSUs

are securely connected by the stable backbone network. TA

can help RSUs to identify the real identity of vehicles on

request.

• Vehicular Nodes: Vehicular nodes are vehicles on the road

which are equipped with an on-board unit (OBU) for

computation and communication, a global positioning system

(GPS) for location service, and an interface for interacting

with drivers.

System Model

• Road Side Units (RSU) and Domain: RSUs are the

infrastructure deployed along the road side which play an

important role in key management, message

authentication/verification, and message dissemination.

• A group of RSUs forms a domain. The number of RSUs

within a domain can be determined based on the

geographical status, infrastructure capacity, deployment plan

and vehicle demography.

• RSUs are further classified into member RSUs (M-RSU) and

leader RSUs (L-RSU).

System Model• Leader Road Side Units (L-RSU): The L-RSUs coordinate with

the trusted authority and generates the group private keys and

group public keys for the vehicles. The L-RSUs also manage and

maintain the database of the group keys. Upon detecting

suspicious behavior, the L-RSUs communicate with the TA to

reveal the identity of the malicious vehicle.

Member Road Side Units (M-RSU): Unlike LRSUs, M-RSUs do

not perform the key generation and management process, but

instead M-RSUs can help vehicles to obtain the group keys from a

leader RSU. Thereby, M-RSUs are semi-trust with the medium

security level. Once the vehicle gets the group key, it can

communicate with any M-RSU inside a domain with the same key.

Proposed Scheme

• The proposed protocol utilizes short group signature

protocol[14] to generate a group private key.

• The leader RSU as a key generator issues group private

keys within a domain.

• In a domain which consists of multiple RSUs, there are one

group public key and many corresponding group private

keys so any member of a domain can sign messages.

• A vehicle can use the same group key with multiple RSUs

within a domain without having to initiate a key

establishment process.

Proposed Scheme

Figure illustrates how vehicles

can request a group private key

to the leader RSU within a

domain.

Secure Key Distribution Scheme

• As a vehicle enters an area of a domain, it can communicate

with any RSU to securely obtain group public/private key pair.

• The secure key distribution scheme is based on the Diffie-

Hellman key agreement protocol for mutual authentication

and sharing a symmetric key.

• Vehicles and M-RSU shares the related parameters to get

the symmetric key.

• gab serves as the secret key KVi_MR between Vi and M-RSU

Secure Key Distribution Scheme

• After establishing a symmetric key, vehicle requests for the

group keys to M-RSU.

• M-RSU forwards the request to the L-RSU.

• L-RSU replies to M-RSU with the group keys for the

vehicle.

• Finally, M-RSU transmits the group keys to vehicle using

the shared symmetric keys.

Secure Key Distribution Scheme

M-RSU

M-RSU

m1

m3m2

m6

m4

m5

L-RSU

Domain

Our Protocol Flow

Regular Communication

Secure Key Distribution Scheme

Evaluation & Analysis

• Simulation Setup

• Manhattan Grid environment simulated in the

Network Simulator

• NS-2 and the mobility simulator SUMO.

• Map of 3600*3600 square meters.

Key Establishment

Key Establishment

• When the domain of multiple RSUs is not

considered, vehicles have to perform the key

exchange procedure with each and every RSUs

separately.

• The figure shows how the average number of key

establishment changes as the vehicles are moving

with/without using domains.

• Here, domain has the area covered by four RSUs

with the vehicles moving randomly.

Key Establishment Delay Comparison

[14]

Key Establishment Comparison

• The computation time of the group signature [14]

measured that the signing and the verification delays

are 3.6 ms and 7.2 ms for key transaction.

• The signing and verification process of the 16 key

establishments without using domain takes

16*(3.6+7.2) ms (172.8 ms).

• On the other hand, while using the domain size of four

RSUs, the time taken for signing and verification of

four key establishments is 4*(3.6+7.2) ms (43.2 ms).

Key Establishment Comparison

• Thus, with the domain size of only four RSUs, the time-

taken for key establishment of the randomly moving

vehicles is 43.2 ms which is 400% more efficient

compared to the 172.8 ms (time taken for key

establishment without domain).

• In practice, the domain size is likely to be higher, so it can

be expected that our scheme will achieve more efficiency.

Group Key Utilization

• Group key utilization time is the time that the vehicle

travels inside the domain after establishing the key.

• Group key utilization time can be used to consider

the frequency of the group key usage in domains and

get idea about average travel time of the vehicles in

various.

• The Figure shows the group key utilization time for

different size of vehicles after receiving the group

keys under the different size of domain.

Group Key Utilization

Group Key Utilization

• It is observed that the vehicles spends around 30-40

seconds in one RSU on average.

• And the average travel time is continuously increasing as

the size of the domain increases.

• When there are four RSUs within a domain, it is

observed that the moving vehicles utilize the group key

about 200% more than the moving vehicles without

having a group key for the domain.

Conclusion• Our scheme provides the scalable solution for the

security of the vehicular networking by using the

concept of domain with multiple RSUs so that a group

key can be utilized for a longer period of time.

• In addition, by splitting the role of the RSU to member

RSU and leader RSU, our approach provides the

distributed key management mechanism.

• Furthermore, our scheme offers the secure key

exchange protocol which allows that group keys can be

securely delivered to vehicular nodes.

References[1] H. Hartenstein and K. Laberteaux, VANET vehicular applications and inter-networking technologies. John Wiley & Sons, 2009, vol. 1.

[2] P. Papadimitratos, A. De La Fortelle, K. Evenssen, R. Brignolo, and S. Cosenza, “Vehicular communication systems: Enabling

technologies, applications, and future outlook on intelligent transportation,” IEEE Communications Magazine, vol. 47, no. 11, 2009.

[3] M. Raya and J.-P. Hubaux, “Securing vehicular ad hoc networks,” Journal of computer security, vol. 15, no. 1, pp. 39–68, 2007.

[4] D. Chaum and E. Van Heyst, “Group signatures,” in Advances in CryptologyEUROCRYPT91. Springer, 1991, pp. 257–265.

[5] Y. Hao, Y. Cheng, C. Zhou, and W. Song, “A distributed key manage- ment framework with cooperative message authentication in vanets,”

IEEE Journal on selected areas in communications, vol. 29, no. 3, pp. 616–629, 2011.

[6] P. Vijayakumar, M. Azees, A. Kannan, and L. J. Deborah, “Dual authentication and key management techniques for secure data trans-

mission in vehicular ad hoc networks,” IEEE Transactions on Intelligent Transportation Systems, vol. 17, no. 4, pp. 1015–1028, 2016.

[7] T. W. Chim, S.-M. Yiu, L. C. Hui, and V. O. Li, “VSPN: VANET- based secure and privacy-preserving navigation,” IEEE Transactions on

Computers, vol. 63, no. 2, pp. 510–524, 2014.

[8] R. Lu, X. Lin, H. Zhu, P.-H. Ho, and X. Shen, “ECPP: efficient conditional privacy preservation protocol for secure vehicular com-

munications,” in Proceeding of IEEE INFOCOM 27th Conference on Computer Communications, 2008. IEEE, 2008, pp. 1229–1237.

[9] W. Diffie and M. Hellman, “New directions in cryptography,” IEEE transactions on Information Theory, vol. 22, no. 6, pp. 644–654, 1976.

[10] M. Gerlach, “Assessing and improving privacy in VANETs,” ESCAR, Embedded Security in Cars, 2006.

[11] J. Freudiger, M. Raya, M. Fe ́legyha ́zi, P. Papadimitratos, and J.-P. Hubaux, “Mix-zones for location privacy in vehicular networks,” in

Proceedings of ACM Workshop on Wireless Networking for Intelligent Transportation Systems (WiN-ITS), no. LCA-CONF-2007-016, 2007.

[12] L. Huang, K. Matsuura, H. Yamane, and K. Sezaki, “Enhancing wireless location privacy using silent period,” in Proceedings of Wireless

Com- munications and Networking Conference, 2005 IEEE, vol. 2. IEEE, 2005, pp. 1187–1192.

[13] K. Sampigethaya, M. Li, L. Huang, and R. Poovendran, “AMOEBA: robust location privacy scheme for VANET,” IEEE Journal on Selected

Areas in Communications, vol. 25, no. 8, 2007.

[14] X. Sun, X. Lin, and P.-H. Ho, “Secure vehicular communications based on group signature and id-based signature scheme,” in

Proceedings of IEEE ICC International Conference, 2007. IEEE, 2007, pp. 1539– 1545.

[15] IEEE Standard 1609.2, “1609.2-2016 - IEEE standard for wireless access in vehicular environments–security services for applications

and management messages,” IEEE Xplore, pp. 1–240, 2016.

[16] S. S. Rosen, “Tamper-proof electronic processing device,” Jul. 11 2000, uS Patent 6,088,797.

[17] K. Lim and D. Manivannan, “An efficient protocol for authenticated and secure message delivery in vehicular ad hoc networks,” Vehicular

Communications, vol. 4, pp. 30–37, 2016.

[18] M. Raya, P. Papadimitratos, V. D. Gligor, and J.-P. Hubaux, “On data-centric trust establishment in ephemeral ad hoc networks,” in

Proceedings of IEEE INFOCOM 27th Conference on Computer Com- munications, 2008. IEEE, 2008, pp. 1238–1246.

•[19] NS-2, “http://nsnam.sourceforge.net/wiki/index.php.”

•[20] simulation of urban mobility, “http://sumo.dlr.de/index.html.”

•[21] I. I. of highway safety, “http://www.iihs.org/iihs/sr/statusreport/article/43/1/1,” Jan 2008.

•[22] R. L. Rivest and A. Shamir, “How to expose an eavesdropper,” Com- munications of the ACM, vol. 27, no. 4, pp. 393–394, 1984.