SecVLC: Secure Visible Light Communication for MilitaryVehicular Networks

Seyhan UcarKoc University,Istanbul Turkey

sucar@ku.edu.tr

Sinem Coleri ErgenKoc University,Istanbul Turkey

sergen@ku.edu.tr

Oznur OzkasapKoc University,Istanbul Turkey

oozkasap@ku.edu.tr

Dobroslav TsonevPureLifi Ltd., Edinburg United Kingdomdobroslav.tsonev@purelifi.com

Harald BurchardtPureLifi Ltd., Edinburg United Kingdomharald.burchardt@purelifi.com

ABSTRACTTechnology coined as the vehicular ad hoc network (VANET)is harmonizing with Intelligent Transportation System (ITS)and Intelligent Traffic System (ITF). An application sce-nario of VANET is the military communication where ve-hicles move as a convoy on roadways, requiring secure andreliable communication. However, utilization of radio fre-quency (RF) communication in VANET limits its usage inmilitary applications, due to the scarce frequency band andits vulnerability to security attacks. Visible Light Communi-cation (VLC) has been recently introduced as a more securealternative, limiting the reception of neighboring nodes withits directional transmission. However, secure vehicular VLCthat ensures confidential data transfer among the participat-ing vehicles, is an open problem. In this paper, we proposea secure military light communication protocol (SecVLC)for enabling efficient and secure data sharing. We use thedirectionality property of VLC to ensure that only targetvehicles participate in the communication. Vehicles use full-duplex communication where infra-red (IR) is utilized toshare a secret key and VLC is used to receive encrypted data.We experimentally demonstrate the suitability of SecVLC inoutdoor scenarios at varying inter-vehicular distances withkey metrics of interest, including the security, data packetdelivery ratio and delay.

Keywordsvehicular ad hoc network; military ad hoc network; visiblelight communication; security

1. INTRODUCTIONTechnology coined as VANET is harmonizing with ITS

and ITF. VANET is proposed to mitigate the problems ofITS and ITF as well as the traffic control and optimization.VANET is a type of ad hoc network that communicates

Permission to make digital or hard copies of all or part of this work for personal orclassroom use is granted without fee provided that copies are not made or distributedfor profit or commercial advantage and that copies bear this notice and the full cita-tion on the first page. Copyrights for components of this work owned by others thanACM must be honored. Abstracting with credit is permitted. To copy otherwise, or re-publish, to post on servers or to redistribute to lists, requires prior specific permissionand/or a fee. Request permissions from permissions@acm.org.

MobiWac’16, November 13-17, 2016, Malta, Maltac⃝ 2016 ACM. ISBN 978-1-4503-4503-3/16/11. . . $15.00

DOI: http://dx.doi.org/10.1145/2989250.2989259

vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I) orboth [1] based on IEEE 802.11p (DSRC), which forms thestandard for Wireless Access for Vehicular Environments(WAVE). One application area of VANET is military ser-vice, namely military ad hoc network. The research in themilitary ad hoc network has gained popularity and expandedto commercial applications that are promising for future.

Figure 1: Military Ad Hoc Network Structure on Highways

In military ad hoc networks, team member vehicles travelas a convoy or along a multi-lane linear road segment inhighway or urban roadways and share data packets witheach other. Fig. 1 demonstrates an example military ad hocnetwork structure on highway in a dashed ellipse that adoptsDSRC as wireless communication technology. On the road-way, military vehicles obey the traffic rules and keep the con-voy structure in order to prevent any attack from adversaryvehicles. During the transportation, tactical command vehi-cle that is the first or the last vehicle in the convoy initiatesdata transmission where the data contains the command orplan and requires timely and reliable delivery [2]. Moreover,the military ad hoc network must ensure that the dissemi-nated data cannot be decoded by other vehicles in the com-munication range when data packets eavesdropped. As aresult, military ad hoc network communication on roadwaysimposes strict requirements on the security of the communi-cation channels used by vehicles and hence requires secureprotocols.

Currently, VANET security solutions mainly focus on thedominant vehicular communication technology, DSRC whichsuffers from the scarcity of RF and it is open to security at-tacks such as jamming and spoofing. Any adversary deviceor vehicle within the transmission range can send the jam-ming signal to block the communication between militaryvehicles. In the spoofing attack, on the other hand, theadversary overhears the DSRC channel and impersonates

Table 1: Comparison of VLC and DSRC key properties

Property VLC DSRCCommunication Scenario Typically LoS Both LoS and NLoSTransmission Range Short Range and Highly Directional Long Range and Usually OmnidirectionalData Rate Up to 400Mb/s Up to 54Mb/sFrequency Band 400 - 790 THz 5.8 - 5.9 GHzPower Consumption Relatively Low MediumSpatial Reuse Efficiency High LowElectromagnetic Interference No YesLicensing Free RequiredCoverage Narrow WideCost Low HighMobility Medium HighWeather Condition Sensitive RobustAmbient Light Sensitive Not Affected

another military vehicle in order to inject faulty informa-tion into a specific area. Although DSRC based VANETtechnologies have evolved over time [3], they suffer from thesecurity vulnerabilities and they are not directly applicableto military communication. One example solution can beenabling the vehicle to have a daily key for communication.However, there exist some incidents where hackers succeedin acquiring the secret key [4] by eavesdropping the DSRCchannel.On the other hand, the scarcity of RF spectrum has led

researchers to investigate alternative technologies. VLC is arelatively new communication technology that uses modu-lated optical radiation in the visible light spectrum to carrydigital information. Recently, many researchers are investi-gating vehicular VLC for different purposes such as channelcharacteristics [5, 6], requirements [7–9] and feasibility in ahybrid architecture with DSRC [10]. Proposed vehicularVLC schemes were studied either experimentally [5, 7–9] orvia computer-based simulations using Lambertian propertyof LEDs [6].One feature that makes VLC superior in comparison to

DSRC is security. The light directivity and impermeabilityof the optical signal facilitate secure data communicationwhere it is ensured that only target vehicles participate inthe communication, making data difficult to receive ratherthan the light coverage. Furthermore, due to the directivityof the VLC transceivers, attackers need to direct strong lightto saturate the receiver which can only be performed on asingle VLC link, as opposed to all vehicles in the communica-tion range in the case of DSRC. From this perspective, VLCis a promising technology to alleviate the security problemsof DSRC in the vehicular environment. However, securityimplication of vehicular VLC had mostly been underratedand there exist only a few studies focusing on non-vehicularscenarios [11, 12]. On the other hand, secure light commu-nication that ensures only the participating vehicles can ex-tract and understand the content of the data is crucial forseveral vehicular applications.In this paper, we propose a secure light communication

protocol (SecVLC) for military ad hoc network on roadwayswhere IR is utilized to share a secret key and VLC is used toreceive encrypted data between vehicles. The contributionof this paper is threefold. First, light directionality propertyis used for ensuring that only target vehicles participate inthe communication. Second, vehicles use full-duplex com-munication where IR is the outgoing link to share a secret

key and VLC is the incoming link to receive encrypted data.We experimentally evaluate the suitability of SecVLC in out-door scenarios at varying inter-vehicular distances with keymetrics of interest, including the security, data packet deliv-ery ratio and delay. Third, to the best of our knowledge, theproposed protocol SecVLC is the first work to secure lightcommunication in the vehicular environment.

The organization of the paper is as follows. Section 2presents the state-of-the-art improvements in VLC technol-ogy. A detailed comparison of DSRC and VLC is demon-strated in Section 3. Section 4 describes the used systemmodel. The details of SecVLC protocol are presented inSection 5, followed by the experimental results in Section 6.Finally, concluding remarks are given in Section 7.

2. VLC TECHNOLOGYThe field of VLC has undergone significant research ad-

vancements over the past decade which only serves to fur-ther emphasize the enormous potential of the technology fora wide range of applications. Gigabit-class connectivity hasbeen demonstrated under laboratory conditions by means ofcommercially available LEDs [13, 14]. Similar speeds havealso been demonstrated with a new class of LED devices,called micro-LEDs, with sizes in the order of tens of mi-crometers. The performance of these micro-LEDs indicatesthat potentially every pixel of every screen, as well as everyindicator light on a device, can be transformed into a high-speed visible light communication transmitter. Further, im-pressive results in the field have shown that laser-based lightsources have the potential to unlock wireless communica-tion rates in the order of hundreds of Gigabits per sec-ond [15, 16]. Complementarily, the high-speed demonstra-tions of the optical transmitters’ capabilities have been witha variety of photodetectors being employed. Avalanche pho-todiodes (APDs) have been widely adopted in high-speedapplications where high sensitivity is required [17], whereassingle photon avalanche photodiode (SPAD) based detectorsare under development and have been demonstrated work-ing with complex modulation schemes such as orthogonalfrequency division multiplexing (OFDM) [18]. In addition,solar panels have also been successfully employed as energy-efficient photodetectors that can provide simultaneous en-ergy harvesting and communication [19, 20]. Consequently,they can be used in a large variety of off-the-grid wirelesscommunication applications. The significant research ad-vancements in the field have been complemented with the

Visible Light

Communication

Infrared

Communication

Source Destination

Source

Destination

Source

Destination

Malicious Actor

Figure 2: System model for military VLC

release of the first LiFi wireless adaptor - the LiFi-X - whichsupports wireless links comparable to existing WiFi net-works, however, in significantly denser deployment scenar-ios [21].

3. COMPARISON OF DSRC AND VLCA comparison of the key properties of VLC and conven-

tional DSRC is presented in Table 1. DSRC is usually omni-directional and can work both in line-of-sight (LoS) and non-line-of-sight (NLoS) scenarios in licensed frequency band 5.8- 5.9 GHz with high mobility. VLC, on the other hand, ishighly directional and typically works in LoS scenarios atshort range, around 25-50 meters, with high sensitivity toweather condition and ambient light. Compared to DSRC,the maximum range of VLC is much shorter as its effec-tive free-space path loss is 4 instead 2 in the case of RF [22].Therefore, VLC provides much higher spatial reuse efficiencywith effective interference control at high vehicle density.Moreover, multipath fading is negligible in VLC even at highvehicle mobility [23]. VLC also brings several advantages ofnot causing any health concern nor any electromagnetic in-terference, being license-free and easy integration with exist-ing LED equipped vehicles with low-cost additional onboardunits. The IEEE 802.15 working group for wireless personalarea networks (WPAN) standardized the PHY and MAClayer for VLC in the IEEE 802.15.7 task group.A typical VLC system uses fast switching light emitting

diodes (LEDs) as the transmitter to simultaneously provideillumination and communication in indoor and outdoor sce-narios. VLC is a promising technology for military ad hocnetwork with most of the communication components al-ready existing within vehicles. Modern vehicles have alreadystarted to use LEDs due to their long service life, high re-sistance to vibration and better safety performance. LEDsare used in the stop lamps, brake lights, turn signals andheadlamps of many vehicles. On the other hand, VLC re-ceivers are mostly either photo-diode (PD) [24] or CMOScamera [25] which can be found in many vehicles as thefront or rear camera for lane tracking and parking purposes.In real-world vehicular VLC deployments, it is difficult

to send messages directly from the front vehicle to all teammembers, which are traveling in a convoy. This is due tothe sharp directivity and the vehicles’ bodies as obstacles.Moreover, vehicles within the coverage of the transmittingvehicle can also eavesdrop the shared packets as in DSRC.From the military ad hoc network perspective, the success-ful decoding of military shared information by an adversarymight have disastrous effects. Therefore, secure vehicular

VLC that ensures only the participating vehicles can extractand understand the content of the data, is required.

4. SYSTEM MODELThe system consists of vehicles moving on the highway

or urban roadways as a convoy, as depicted in Fig. 2. Inthe vehicular convoy, vehicles are organized into groups ofclosely following vehicles and obey the traffic rules such asspeed limit and usage of headlights. The speed variationof the vehicles in the convoy is very small. Each vehiclemaintains a proper following distance by either slowing downwhen it gets too close or speeding up to preserve the convoydistance that is minimum 2 meters [26]. Moreover, theremay exist a malicious insider in the system that overhearsthe communication channel and tries to extract the datacontent. The malicious insider can be a roadside unit or avehicle that is not part of military ad hoc network.

The message is broadcast in the vehicle convoy in multiplehops. In other words, each vehicle can be both source anddestination as shown in Fig. 2. In each hop, the commu-nication of consecutive vehicles is provided via VLC, sincefurther nodes cannot communicate with other vehicles in be-tween. Each vehicle contains a transmitter unit connectedto the LED headlights and IR receiver on the front bumper.The data is disseminated through these headlights from sourceto destination.

Vehicles are also equipped with the PD based receiver unit(Rx) and IR transmitter on the rear bumpers. Transmitteddata packets are captured by Rx whereas the secret key istransmitted from destination to source by use of IR trans-mitter. Each vehicle uses IR as the outgoing link to sharea secret key and VLC as the incoming link to receive en-crypted data. To ensure shared secret key freshness, it ischanged for each data transmission in order to prevent thesystem from a possible attack that can be triggered by themalicious insider. Due to LoS sensitivity of IR, it is onlyused for 4-bytes secret key dissemination.

5. SECVLC PROTOCOLFeatures of the proposed secure light communication pro-

tocol SecVLC are as follows;

1. It uses the directionality property of VLC to ensureonly target vehicles participate in the communication.

2. It utilizes the full-duplex communication where IR isthe outgoing link to share a secret key and VLC is theincoming link to receive encrypted vehicle data.

Source Destination

Initialize()

GenerateKey()

Key GenerationKey Generation

Secret Key

Data Encryption()

EncryptionEncryption

Encrypted Data Packet

Data Decryption()

DecryptionDecryption

Secret Key

SecVLC ProtocolSecVLC Protocol

Figure 3: SecVLC Protocol steps

3. It operates with keys generation and share mechanismthat is used for the data encryption and decryptionwhere data packets cannot be decrypted without gen-erated keys.

Fig. 3 demonstrates the steps of SecVLC protocol. SecVLCconsists of five parts; initializing system, key generation, IRkey transmission, data encryption/decryption and VLC en-crypted data dissemination. SecVLC starts by initializingthe system. The system boots up its hardware componentsand informs the destination that it is waiting for the secretkey. The destination is triggered via the incident light beamcoming from the source. When the destination receives thelight beams then it generates a secret key for data encryp-tion.Generated secret key is transmitted via IR transmitter.

The narrow transmission angle property of IR enables onlythe following vehicle to receive the secret key, as opposed toall vehicles in the communication range in the case of DSRC.Secret keys are based on the Advanced Encryption Standard(AES) that is widely adopted due to features such as fastsymmetric key generation and strength compared to other

alternatives without any practical attacks against AES tillto date.

After receiving the secret key from the destination, thesource encrypts the data packet and transmits the encryptedpacket via light beams in VLC. If any vehicle exists in thelight coverage of the source, it cannot decode the data packetwithout the secret key. This actually solves the channel over-hearing problem of DSRC based military communication.

Following the sharing of the secret key, the destination re-ceives encrypted data packets from the source while concur-rently sharing newly generated secret key via IR transmitterfor the next round of data transmission. The full-duplex IRand VLC communication enable the data security withoutincreasing delay. After receiving data packets, the desti-nation decrypts them using the secret key. For each datamessage, the destination shares a secret key with the sourcefor encryption.

6. PERFORMANCE EVALUATIONWe implemented SecVLC protocol in Java on top of Li-1st

transceiver software [27] that is integrated Keyczar [28] keygeneration toolkit. Li-1st is the first commercial product ofVLC that is manufactured by pureLifi Ltd. It provides anopportunity to rapidly develop and test VLC applicationsthat utilize commercial LED infrastructures. Li-1st consistsof transmitter unit Tx and PD based receiver unit Rx. TheTx unit is attached to two symmetrical LED fog lights [29]where automotive fog lights are preferred due to their wideand flat illumination pattern to minimize reflection by fog.On the other hand, Keyczar is an open source toolkit devel-oped by Google for key generation.

VLC Tx

VLC Tx VLC Rx

IR RxIR Tx

Figure 4: VLC and SecVLC Experimental Setup

Two Vishay [30] high speed infra-red emitting diodes areutilized as IR transmitter and IR receiver for sharing thesecret key between source and destination. Both Tx and Rxare connected to computers for evaluating communicationperformance. In order to compare the security vulnerabili-ties of communication medium, scenarios where vehicles useDSRC and visible light data transmission, namely VLC, areevaluated. The DSRC communication scenario is simulatedwith the convoy driving implemented simulator, VEhicularNeTwork Open Simulator (VENTOS) [31]. On the otherhand, VLC and SecVLC experiments are performed in anoutdoor environment as shown in Fig. 4 to take into ac-count the reflections from vehicles and road. Night time out-door measurements are executed to compensate shot noise,sourced by diurnal variations. Our experiment emulates thescenarios that are the front of following vehicle disseminat-ing commands (i.e. mission orders, mission plan and etc.)with LED fog lights to the rear of leading vehicle proceed-

ing on a curved path. Table 2 lists the experimental systemparameters.

Table 2: Experimental Setup Parameters

Parameters ValueLED Fog Lights Ground Height 36 cmLED Fog Lights Separation Distance 150 cmInter-Vehicular Distance 2 - 6 metersData Packet Size 100 bytesLi-1st Modulation Pulse Amplitude

ModulationLi-1st Error Correction Reed-SolomonLi-1st Data Rate 5 MbpsSecret Key Packet Size 4 bytesVishay IR Half Intensity 18 degree

Performance evaluation of SecVLC is done in two parts.The first part focuses on the security analysis of SecVLCwhere the system with a malicious insider is investigated.The malicious insider is a vehicle that is positioned on theroad with constant mobility. For each experiment, 100 datapackets are sent over the military ad hoc network and mali-cious insider tries to extract the data content. In the secondpart of the performance evaluation, network performancemetrics are interpreted by comparing the data packet deliv-ery ratio (DPDR) and the delay between vehicles. In eachexperiment, custom created 100-bytes data packet is sent.The effect of data size is considered for the different volumeof data varying from 200 bytes to 500 bytes.

6.1 Security AnalysisIn security analysis of SecVLC protocol, malicious vehi-

cle’s data decoding ratio is analyzed. Data decoding ratio isdefined as the ratio of the number of successfully plain textconverted data packets to the total number of transmitteddata packets. In this scenario, the malicious vehicle receivesthe data packets and tries to decode the data for subsequentprocesses such as stealing the vehicle identity information.

2 3 4 5 6

0

10

20

30

40

50

60

70

80

90

100

Distance (meters)

Data

Packet

DecodingRate

(%)

DSRC

VLC

SecVLC

Figure 5: Data Packet Decoding Ratio Comparison

Fig. 5 demonstrates that adversary vehicle can receive thedata packet in both DSRC and VLC scenarios with mini-mum %70 data packet decoding ratio. In the DSRC, ad-

versary vehicle overhears the channel if it is located in thetransmission range (300 meters) of military vehicles. Onthe other hand, VLC limits the adversary data receptiondue to its directional transmission. However, adversary ve-hicle still receives the data if it is positioned in headlightcoverage. Compared to DSRC and VLC, SecVLC encryptsthe data packet and data content can only be decryptedwith the secret key. Even if the adversary vehicle overhearsthe channel, it can only receive plain text control packetstransmitted in the initialization phase of the protocol.

6.2 Network Performance AnalysisIn this part of the performance evaluation, network perfor-

mance of SecVLC protocol is investigated by analyzing themetrics including DPDR and the delay. DPDR is defined asthe ratio of the number of successfully received data pack-ets to the total number of transmitted data packets. Theaverage delay metric is defined as the average latency ofdata packets that travel from the Source to the Destina-tion that includes secret key IR transmission, data encryp-tion/decryption and VLC dissemination.

2 3 4 5 680

82

84

86

88

90

92

94

96

98

100

Distance (meters)

Data

Packet

DeliveryRatio(D

PDR)%

VLC 200 Byte

SecVLC 200 Byte

VLC 400 Byte

SecVLC 400 Byte

VLC 500 Byte

SecVLC 500 Byte

Figure 6: Data Packet Delivery Ratio Comparison

Fig. 6 shows the DPDR comparison of SecVLC and VLCat different distances for varying data packet size. We ob-serve that the DPDR value exhibits similar degradation pat-terns with the increasing distance. Moreover, as the dis-tance gets larger, both SecVLC and VLC have difficultyin delivering data packets. This can be explained by thereceived signal strength (RSS), where as the distance getslarger the RSS sensed in receiver unit decreases. As a resultof RSS decrease, data packets cannot be received success-fully. From this perspective, we can say that RSS decreasein the large distances is the major factor that affects theDPDR in SecVLC.

Fig. 7 shows the average delay performance of SecVLCprotocol compared to VLC as a function of distance withvarying data size. Compared to VLC, measured delay valuefor SecVLC contains key IR transmission, data encryption,data decryption and VLC data dissemination. Moreover,the effect of data size on average delay is analyzed by chang-ing the data volume. As observed in Fig. 7, as the data sizeincreases SecVLC necessitates larger time to encrypt and de-crypt. Despite the average delay for SecVLC is higher than

2 3 4 5 60.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

Distance (meters)

AverageDelay

(sec)

VLC 200 Byte

SecVLC 200 Byte

VLC 400 Byte

SecVLC 400 Byte

VLC 500 Byte

SecVLC 500 Byte

Figure 7: Average Delay Comparison

the VLC, it is acceptable provided that secure data trans-mission is enabled where only target vehicle can extract thedata content. From this perspective, we can say that there isa trade-off between security and delay for VLC and SecVLCprotocol: VLC provides lower delay than SecVLC whereasSecVLC achieves data security by encrypted communica-tion.

7. CONCLUSION AND FUTURE WORKIn this study, we perform the first work to investigate data

security in light based military ad hoc network and proposeSecVLC protocol for securing communication. In SecVLC,we first use the VLC directionality property to ensure onlytarget vehicles participate in the communication. Then, ve-hicles use full-duplex communication where IR is the out-going link to share a secret key and VLC is an incominglink to receive encrypted vehicle data. We experimentallyevaluate the suitability of SecVLC in outdoor scenarios byvarying the inter-vehicular distance and data size with dif-ferent metrics of interest including the security, data packetdelivery ratio and delay.Experimental evaluation of SecVLC protocol demonstrates

its suitability for securing light based military communica-tion. In the security analysis of SecVLC, we observe thatdespite VLC limits the data reception due to its directionaltransmission, it still possible to receive and decode the datapacket if the adversary locates in light coverage. On theother hand, secret key enabled SecVLC prevents adversaryvehicle decoding the data packet even it is received success-fully. Moreover, the network performance analysis showsthat DPDR value exhibits similar degradation patterns withthe increasing distance for both SecVLC and VLC. RSS atlarge distance is not enough to successfully receive the datapacket and it is the major factor that has an effect on theDPDR in SecVLC. On the other hand, delay value analysisshows that SecVLC requires extra time for data encryptionand decryption. As the data packet size increases, delayvalue also increases.In the future, we plan to extend SecVLC protocol by us-

ing VLC for both key and data transmission such that ex-tra IR transceiver will not be required by the protocol. Weaim at determining the efficiency of SecVLC protocol in LoS

scenarios and analyzing the performance of NLoS communi-cation where the vehicle does not have a direct view of therear vehicle bumper. We will also experimentally evaluatethe SecVLC protocol at different light, temperature and hu-midity conditions for high-speed data communication withautomotive LED lights.

AcknowledgementOur work was supported by Argela and Turk Telekom un-der Grant Number 11315-07. Sinem Coleri Ergen acknowl-edges support from Bilim Akademisi - The Science Academy,Turkey under the BAGEP program, and the Turkish Academyof Sciences (TUBA) within the Young Scientist Award Pro-gram (GEBIP). Moreover, authors also acknowledge the tech-nical and hardware support from the pureLiFi Ltd.

8. REFERENCES[1] S. Ucar, S. C. Ergen, and O. Ozkasap,

“Multihop-Cluster-Based IEEE 802.11p and LTEHybrid Architecture for VANET Safety MessageDissemination,” IEEE Transactions on VehicularTechnology, vol. 65, pp. 2621–2636, April 2016.

[2] I. Rubin, A. Baiocchi, F. Cuomo, and P. Salvo,“Vehicular Backbone Network Approach to VehicularMilitary Ad Hoc Networks,” in MilitaryCommunications Conference, MILCOM IEEE, Nov2013.

[3] R. G. Engoulou, M. Bellaiche, S. Pierre, andA. Quintero, “VANET security surveys,”ComputerCommunications, 2014.

[4] “Hacking The Vehicles By Listening Medium.”http://goo.gl/3aGHwr.

[5] W. Viriyasitavat, S.-H. Yu, and H.-M. Tsai, “Shortpaper: Channel model for visible light communicationsusing off-the-shelf scooter taillight,” in VehicularNetworking Conference (VNC), IEEE, Dec 2013.

[6] P. Luo, Z. Ghassemlooy, H. L. Minh, E. Bentley,A. Burton, and X. Tang, “Performance analysis of acar-to-car visible light communication system,”Appl.Opt., Mar 2015.

[7] S. Ucar, B. Turan, S. Coleri Ergen, O. Ozkasap, andM. Ergen, “Dimming Support For Visible LightCommunication in Intelligient Transportation andTraffic System,” in Urban Mobility and IntelligientTransportation System (UMITS), IEEE/IFIP NetworkOperations and Management Symposium (NOMS),2016.

[8] B. Turan, S. Ucar, S. Coleri Ergen, and O. Ozkasap,“Dual Channel Visible Light Communications forEnhanced Vehicular Connectivity,”VehicularNetworking Conference (VNC), IEEE, 2015.

[9] H.-Y. Tseng, Y.-L. Wei, A.-L. Chen, H.-P. Wu,H. Hsu, and H.-M. Tsai, “Characterizing linkasymmetry in vehicle-to-vehicle Visible LightCommunications,” in Vehicular NetworkingConference (VNC), IEEE, Dec 2015.

[10] S. Ishihara, R. V. Rabsatt, and M. Gerla, “ImprovingReliability of Platooning Control Messages UsingRadio and Visible Light Hybrid Communication,”Vehicular Networking Conference (VNC), IEEE, 2015.

[11] A. Mostafa and L. Lampe, “Physical-layer security forindoor visible light communications,” in

Communications (ICC), IEEE InternationalConference on, June 2014.

[12] B. Zhang, K. Ren, G. Xing, X. Fu, and C. Wang,“SBVLC: Secure barcode-based visible lightcommunication for smartphones,” in INFOCOM,Proceedings IEEE, April 2014.

[13] A. M. Khalid, G. Cossu, R. Corsini, P. Choudhury,and E. Ciaramella, “1-gb/s transmission over aphosphorescent white led by using rate-adaptivediscrete multitone modulation,” IEEE PhotonicsJournal, Oct 2012.

[14] G. Cossu, A. M. Khalid, P. Choudhury, R. Corsini,and E. Ciaramella, “2.1 gbit/s visible optical wirelesstransmission,” in Optical Communications (ECOC),38th European Conference and Exhibition on, Sept2012.

[15] D. Tsonev, S. Videv, and H. Haas, “Towards a 100Gb/s visible light wireless access network,”Opt.Express, 2015.

[16] A. Gomez, K. Shi, C. Quintana, M. Sato, G. Faulkner,B. C. Thomsen, and D. O’Brien, “Beyond 100-Gb/sIndoor Wide Field-of-View Optical WirelessCommunications,” IEEE Photonics TechnologyLetters, Feb 2015.

[17] Y. Wang, X. Huang, L. Tao, and N. Chi, “1.8-Gb/sWDM visible light communication over 50-meteroutdoor free space transmission employing CAPmodulation and receiver diversity technology,” inOptical Fiber Communications Conference andExhibition (OFC), March 2015.

[18] O. Almer, D. Tsonev, N. A. W. Dutton, T. A. Abbas,S. Videv, S. Gnecchi, H. Haas, and R. K. Henderson,“A SPAD-Based Visible Light CommunicationsReceiver Employing Higher Order Modulation,” inIEEE Global Communications Conference(GLOBECOM), Dec 2015.

[19] Z. Wang, D. Tsonev, S. Videv, and H. Haas, “On theDesign of a Solar-Panel Receiver for Optical Wireless

Communications With Simultaneous EnergyHarvesting,” IEEE Journal on Selected Areas inCommunications, Aug 2015.

[20] S. Zhang, D. Tsonev, S. Videv, S. Ghosh, G. A.Turnbull, I. D. W. Samuel, and H. Haas, “Organicsolar cells as high-speed data detectors for visible lightcommunication,”Optical, Jul 2015.

[21] “LiFiX.” http://purelifi.com/lifi-products/lifi-x/.

[22] T. Komine and M. Nakagawa, “Fundamental analysisfor visible-light communication system using LEDlights,” IEEE Transactions on Consumer Electronics,vol. 50, Feb 2004.

[23] Z. Cui, C. Wang, and H. M. Tsai, “Characterizingchannel fading in vehicular visible lightcommunications with video data,” in VehicularNetworking Conference (VNC), IEEE, 2014.

[24] S.-H. Yu, O. Shih, H.-M. Tsai, N. Wisitpongphan, andR. Roberts, “Smart automotive lighting for vehiclesafety,”Communications Magazine, IEEE, December2013.

[25] T. Yamazato, I. Takai, H. Okada, T. Fujii, T. Yendo,S. Arai, M. Andoh, T. Harada, K. Yasutomi,K. Kagawa, and S. Kawahito, “Image-sensor-basedvisible light communication for automotiveapplications,”Communications Magazine, IEEE, July2014.

[26] M. Amoozadeh, H. Deng, C.-N. Chuah, H. M. Zhang,and D. Ghosal, “Platoon management withcooperative adaptive cruise control enabled byVANET,”Vehicular Communications, 2015.

[27] “Li-1st.” http://purelifi.com/li-fire/li-1st/.

[28] “Google Keyczar Tool.” https://goo.gl/LMuY1l.

[29] “LEDFog101 OSRAM.” http://goo.gl/ty5zEC.

[30] “Vishay Infra-red (IR) Diodes.” http://goo.gl/2DIKcq.

[31] “VEhicular NeTwork Open Simulator (VENTOS).”http://goo.gl/OueFkO.