WCNC 2012 Workshop on Wireless Vehicular Communications and Networks

Getting the message about road safety

Chris Wright Zory Marantz Penina Orenstein
Emeritus Professor of Transsport Management NYC College of Technology Seton Hall University
Middlesex University, London, UK Brooklyn, New York 11201, USA South Orange, NJ 07079, USA
Email: christpherwright@btinternet.com Email: zory@ieee.org E-mail: orenstpe@shu.edu

Abstract—This paper reviews a range of problems in the road a focal point on the main uses of this technology to better
transport field and the potential role of vehicular ad-hoc network choose the terminology that we’ll propose for it.
systems (VANETs) in helping to solve them. In reality, the
communications requirements vary widely from one application A. Different types of application
to the next, in terms of range, latency, and connectivity together
with vehicle and roadside hardware. Based on the nature of the Reference [3] distinguishes broadly between active safety
wireless channel, this presents some challenges for communica- and business/entertainment, and advocate a common approach
tion to security. The paper concludes with a summary of the bridging across both, as do [4] in their summary of the NoW
current state of VANET technology and presents a summary of project. Reference [5] further divides applications notionally
the challenges to be found in each approach.
Index Terms—vehicular, networks, mobile, ad-hoc, VANET into 4 groups: Active safety, Public service, Improved driving,
and Business/entertainment (infotainment)
I. I NTRODUCTION However, several authors distinguish between hazards that
Much of the early research into VANET applications was demand immediate action, as opposed to hazards of which the
driven by German car manufacturers. The FleetNet project, driver merely needs to be aware. For example, [6] propose
which started in 2001, postulated a vision for car-to-car different strategies for handling ‘Safety of life’ and ‘Safety’
communications as a huge MANET running mainly internet messages, although in their case the distinction is confusing
applications. The outcome was a communications platform because the latter term seems to refer to routine beacon-
configured to perform multi-hop unicast packet-forwarding. ing messages that other authors would not regard as safety
Currently, the technology that would be used on a large scale messages as such, but rather general-purpose messages that
VANET network for practical purposes is being rolled out in support a wide range of applications.
today’s modern vehicles in the form of sensors [1], [2]. This is Reference [5] graduates the hazard scale differently: a
being done mostly for safety purposes in avoiding automobile potential crash is treated as a sequence of stages leading up to
accidents and automatic parking. The result of this first step impact, and different applications are classified according to
is that we are currently on the way to properly implementing the stage at which they are expected to intervene (Dangerous
a complete intelligent transportation system (ITS). road features, Abnormal traffic and road conditions, Danger of
None the less, with all the research that is being done, collision, Crash imminent, Incident occurred). [7] distinguish
there is still a lack of consistency in the use of terminology between ‘situations’ and ‘events’, and analyze several factors
that the authors feel are hindering the rate of development that can influence the reliability of safety information and the
of the technology and its rapid deployment. In addition, we effectiveness of the applications that may depend on it.
investigate the many approaches and systems that have been B. Individual safety applications
created in achieving the ITS vision.
The rest of this paper is organized as follows. Section II Car manufacturers have identified many scenarios, each
will present the various types of communications that are reflecting a specific hazard or accident type. A total of 16
possible within a VANET environment. In Section III we applications are listed by [7]. At the same time, the US
resent the various ways that information my be processed CAMP project assessed 34 safety applications and mentioned
within or external to the VANET. Section IV presents the several others. Eight were chosen for more detailed analysis
various patterns of communication and how the information as representing the likely range of communication demands
makes its way from point source to destination. Section V made on a VANET system. [8] More recently, [9] highlight
shows the security breaches that are possible within a VANET four basic areas concerning highway safety: Pre-crash sensing,
utilizing today’s current technologies. Section VI presents the Cooperative collision warning, Emergency electronic brake
future challenges that have yet to be met. Finally, the paper lights, and a Lane change supervisor.
concludes with a summary. Only the last three are common to all the publications cited.
Hazards are broken down into narrowly defined categories, and
II. A PPLICATIONS applications are conceived to deal with them individually. Two
VANET technology has many different applications besides possible exceptions are the ‘Virtual warning signs’ put forward
for safety and entertainment. We mention them to allow us by [10] and [5], and the Cooperative Intersection Collision

978-1-4673-0682-9/12/$31.00 ©2012 IEEE 322

Avoidance System (CICAS) featured in [11] together with A. In-network processing
‘Vehicle-to-vehicle anti-crash warning’ that appears in [7]. While some authors see regular beaconing as an auxiliary
function essential to hazard warning applications [31], others
C. Non-safety applications see it as an application in its own right. A receiving vehicle
The CAMP project [8] report assessed eleven ‘non-safety’ (call it vehicle B) keeps track of the messages by storing
applications and mentioned several others. [12] regard non- their contents in a location table [6], otherwise known as
safety applications as the key to market penetration. Other the ‘neighbor table’ [14]. From the contents of its table at
interesting proposals featured in other publications are Traffic any particular moment, vehicle B can infer the speeds and
jam detection [13] and Detecting vacant parking spaces [14], positions of the vehicles closest to it, in other words it can
[15]. The previous applications make the distinct assumption identify its nearest neighbors and track their movements.
that the vehicle(s) have human control. There is however If vehicle B deduces that it is on a collision course with
another application of full vehicle automation. [16]–[18] Be- vehicle A, its on-board unit can be programmed to issue an
fore delay sensitive information is used for safety purposes audible warning. In addition, vehicle A can broadcast threshold
on a VANET, it is these non-safety applications that will be information from its sensors. To which Vehicle B would react
the gateway tests to prepare VANETs for more processing accordingly. Such categories of information are richer but less
and communication intensive applications such as cooperative objectively reliable. Accordingly, any vehicle B that receives
collision mentioned in Section II-B. the beacon can add information to it before passing it on to
its neighbors to notify them of an imminent accident about
D. Performance requirements to occur. Irrelevant information is dropped. In this scenario,
Wireless radio technology by itself does not make the road a broadcast messages may be propagated from vehicle to vehicle
safer place, but it does enable a range of applications that rely over large distances, at the cost of some delay at each node, a
on communications. In particular, it can be used to measure process that has been described as single-hop with in-network
separation between vehicles equipped with accurate GPS units, processing [4], [21], [31]. It allows vehicles to collaborate by
by exchanging location data. Wireless nodes are sometimes evaluating, sharing and filtering information in such a way as
densely packed, leading to potential ‘broadcast storms’ and to improve reliability and at the same time, reduce the number
interference, while at the other extreme, vehicles may be too of transmissions and hence the load on the wireless network.
far apart to communicate at all [4]–[6], [19]. However, information needs to be interpreted in context [3].
For safety applications, where events on the road can change The problem is expressed in a more direct way by [7]. The
in a fraction of a second, fast communication is vital. A evaluation of incoming safety messages may be affected by
maximum latency figure of 100 ms has been cited as a basic circumstances, a phenomenon they call ‘situation dependency’.
requirement [8], [20]. Non-safety applications are less de- Important sources of variation are (a) driver-related determin-
manding, but they do have specific requirements in economic, ism and (b) configuration dependency. Judging the implica-
functional, performance and deployability terms [21]. tions of (a) or (b) is difficult, especially in a continuously
Most of the research done in VANET works with off-the- changing situation. Events that can be detected objectively
shelf technology, specifically the IEEE 802.11p standard. [22] and deterministically are the most promising candidates for
Recently there have been investigations in utilizing multiple VANET safety applications.
varying technologies such as GPS, 802.11, LTE , and on-board Some work has been done for specific situations such as
sensor technology to achieve the results necessary for efficient traffic light and intersection control. [32]–[34] The control is
communications. [23]–[25] Since 802.11 was designed with a done not only for safety but also for efficiency, specifically
base station in mind, the protocols used for medium access optimizing fuel economy.
result in various issues such as broadcast storms and hidden B. Outcome and response
nodes which are active areas of research. [26]–[30] In that The alternative responses generated by a VANET safety
research there seem to be different sets of terminologies system can be classified into three groups [7]: autonomous
compared to what the VANET society is using. action, e.g. the system applies the vehicle brakes without
reference to the driver, a warning is issued to the driver that
III. I NFORMATION PROCESSING
immediate action is required, and an ‘awareness’ alert is issued
Information processing is a major consideration when deal- to the driver. Almost all applications work in the same way, by
ing with safety applications. The information obtained from producing a computer-generated alert to spur the driver into
the environment before, during, or after an incident needs to doing something (or not doing it as the case may be).
be processed and a decision made whether to send out any The ergonomics of the driver interface must be considered.
notifications of the incident. These notifications may be sent If only a small proportion of these applications appeared on the
to the surrounding vehicles and processed by them explicitly. market, drivers could be inundated with alerts. Experience in
The other option is to have the nearest neighbors closest to the airline industry suggests that repetitive alerts are ignored,
the scene of the event pass on any information that may be and in an emergency, they can be distracting and even counter-
relevant to the proper authorities. productive. [35]

323
 Some of the literature tends to provide autonomous systems and modified according to context where appropriate before
that take over control of the vehicle to achieve the desired any information is passed on. As a result, duplicate or redun-
objective performance. [32], [33] [34] provide a model where dant information is systematically discarded, which helps to
the driver is advised a specific speed, but considers the reduce pressure on the wireless network and avoid potential
situation when the driver does not heed the suggestion and ‘broadcast storms.’
stops the car automatically if necessary.
B. Store-and-forward
IV. PATTERNS OF COMMUNICATION
Store and forward is an essential part of any VANET system
Road traffic can be regarded as a system in which not because at any time traffic density can fall below the threshold
only drivers but also highway authorities and commercial level necessary to sustain wireless connections. [4], [5], [7] In
agencies participate. Different VANET applications call for sparse traffic conditions, it is necessary for vehicles to hold
radio messages to be instigated by and delivered to different and re-broadcast a hazard message at intervals to ensure that
groups of participants in different ways: beaconing, unicas- it is passed on. Such messages will need to be accompanied
ting, multicasting, etc. An example of different applications by ‘Time of validity’ and ‘Area of validity’ information [6]
matched with different communication schemes appears in [5]. so they do not bounce around the road network indefinitely.
In particular, safety applications and non-safety applications Alternatively, roadside units could be used to store and pass on
tend to require quite distinct patterns of communication. Like messages to vehicles approaching remote locations, but [36]
any message sent over the internet, a non-safety message is cite three reasons why store-and-forward might not be easy to
composed of data packets [21] that are independently directed implement in practice, the main reason being delays introduced
towards a target that could be several kilometers away, but the by edge nodes that would have to process all the packets in
target has a known address. The data packets are routed and the queue and then transmit them causing a bottleneck.
delivered without reference to their contents. In other words,
non-safety messages are unicast multi-hop [4]. C. Interference, redundancy and overload
By contrast, safety messages are warnings delivered either
Although wireless radio signals do not require clear line of
to vehicles that happen to be nearby or vehicles in a defined
sight, they are subject to degradation from various sources.
geographical area. It is the relative position of the ‘target’
Interference is inevitable since the VANET signals broadcast
vehicle that matters, not who is driving it. Several papers
by neighboring vehicles share the same frequency. [36], [37]
refer to the gathering of ‘state information’ with ‘processing
The likelihood of a broadcast storm increases with the size of
at network level’, the information being ‘consumed where it
road network. It is necessary to eliminate duplicate messages
is generated.’ [21]
where possible, and this is partly the motivation for in-network
Awareness beaconing is a safety message that can be
processing [7], but of course there is then the trade-off with
pictured as a short-range status message broadcast at frequent
delay, as mentioned in Section IV-B
and regular intervals to any vehicle that happens to be in
radio range. [4], [14] Such messages are not addressed to any Another reduction method that has been taking center stage
particular vehicle and recipients do not acknowledge receipt, is that of cohorts of vehicles within a specific piece of roadway
so the sender cannot be sure whether anyone is taking any as a result of natural patterns that result under normal traffic
notice. conditions. [18], [38]–[41] This perspective of group dynamics
Unicast and multicast, on the other hand, refer to messages is useful not only in information reduction, but also in the
intended for pre-identified targets, and they are used mainly use of other tools such as game theory to facilitate efficient
for non-safety applications. These are the modes that would be performance within certain parameters acceptable to all group
used for downloading/uploading information from neighbors members. [42], [43]
or utilizing the neighbors as relays for communication with an
V. S ECURITY
infrastructure backbone.
To preserve privacy, a vehicle must not have a globally
A. Routing and forwarding unique recognizable permanent identifier [7]. Among others,
In mobile ad-hoc network (MANET) and non-safety com- [4] have recommended encryption together with frequent inter-
munications in VANETS, messages handling is organized change of pseudonyms which triggers substitution of addresses
around the data packet as the unit of communication. [21] on all protocol layers in a node. The pseudonymity con-
Each packet has a specific destination, and is relayed to cept, certification authority, pseudonymous authentication, and
its destination intact irrespective of its content or meaning, signed beacons are all discussed in more detail by [20], who
an approach referred to as ‘address-centric routing’ [31], or conclude that available encryption techniques push existing
alternatively ‘Packet-Centered Forwarding’ (PCF) [6]. on-board processors to their limits. But by far the biggest
By contrast, in ‘data-centric routing’ [31], also referred to threat to a VANET system is jamming of the GPS radio
as ‘Information-Centric Forwarding (ICF)’ [6], the forwarding signals, whose signal is necessary to a VANET for knowing
process takes into account whatever information the data is the locations of its members. This can be done with a device
intended to convey. At each node, the contents are evaluated being made in China and sold for $30 [44].

324
 None the less, there have been great strides in security in II, [59]–[62] go through the most recent implementations and
terms of implementation and complexity. [45] propose a way managements of VANET systems including using commercial
to look at the various levels of abstraction of security depend- products such as smart phones. That is a positive step forward
ing on which social body is involved, i.e. academia, industry, since consumers are already familiar with the technology and
government, and end users. As in previous section [46] investi- will be less hesitant in accepting it. Hence, one of the things
gate how communication between cohorts may be kept secure. that must be considered is the necessary infrastructure that
[47] propose a scheme that attempts to guarantee maximum must me set up in order to achieve a high level of market
security in terms of authentication, anonymity, unlinkability, penetration.
traceability, integrity, non-repudiation, vehicle revocation, and The key issue here is the comfort level that the consumers
resisting replay attack. This is a difficult task since many and drivers will have with the automated system by giving up
of the characteristics are opposites of each other. This may control. Several laws internationally require a human driver
only be achieved with the help of third-party involvement as to be in control of motorized vehicles or be able to take
in the In addition to covering the major areas of security, over instantly. If consumers do not feel comfortable and the
[47] also shows great efficiency improvements in terms of insurance industry does not feel safe about these safety and
the time it is necessary to process each of the individual infotainment systems being installed, there may be no future.
security modes. However, as was mentioned in SectionIV-C, In fact, the incremental changes in the technology can only go
there are opportunities to implement other alternative security so far until it becomes necessary to make a huge investment
implementations methods. For example, [42] introduce fuzzy in infrastructure to truly reap the rewards of VANETs. [63]
logic and fictitious play to deploy countermeasures in the event
of an attack such as jamming. Finally, it is imperative to VII. C ONCLUSIONS
consider the overhead generated due to security protocols. [48] In a conventional network, the endpoints of a message are
defined in advance. The originator and recipient(s) of unicast
VI. F UTURE RESEARCH CHALLENGES and multicast messages have fixed digital addresses, so that
There are very fundamental impediments that are standing the communication layers in the protocol stack can handle
in the way of mass acceptance and implementation of VANET dissemination independently of the contents of the message,
technology. One of them being the constant modeling that but safety applications are different. The OBU has no idea
is done to obtain better simulations of the various proposed who to talk to. It must first identify a target.
systems and algorithms. Even with the 802.11p and 1609.x Awareness beaconing makes practical sense. Safety mes-
standards, collectively known as wireless access for vehicular sages are targeted towards vehicles further away, but the
environments (WAVE) [49], there are still many papers on target is a geographical area rather than a set of addresses.
modeling the VANET environment Models ranging from accu- The intended destination, mode of forwarding and message
rate models of signal propagation [50]–[53] to mobility models content are all determined by the safety application software
[34], [54]–[56]. Most of the research is done in a highway rather than the communications software. And if the message
environment, not considering pedestrians, cyclists, animals, is determined by the application software, it follows that
busses, trucks, etc. In urban and sub-urban environments, the applications software and communications software must
these entities must also be considered within an accurate interact somehow. The implication is that a cross-layered
environmental model and rural environment also need to be solution is necessary.
tested. Moreover, information is subject to uncertainty. The relia-
Also, given the amount of communication that becomes bility of a warning message to a driver can be improved by
realized with a VANET system, issues concerning spectral combining information from several different vehicles and/or
utilization are impossible to ignore. In addition to the channel running algorithms that are capable of working with such
modeling just mentioned, there are steps being taken to im- uncertainties such as fuzzy logic and fictitious games. This
prove spectral efficiency. [57], [58] Also, the various frequency implies a degree of ‘in-network processing’ or cooperation.
ranges being used by different systems, i.e. 5.9 GHz in the Not only does this improve reliability, it also reduces the
US and 700 MHz in Japan, imply that a ubiquitous system communications load on the network, which can easily be
be used for simple implementation within any current or overwhelmed in a ‘broadcast storm’ when the road is con-
future region that would adopt a VANET system. The device gested.
that will become the most pervasive will be the software These complications may explain why in Europe, research
defined radio (SDR). With the numerous technologies being has moved away from multi-hop IP towards the single-hop
implemented within VANET systems, as mentioned in Section model that has dominated throughout in the US and Japan.
II-D, an SDR will be the best method for implementing the But there are misgivings. [36] claim that single hop is not
necessary versatility in communications whether for safety or sufficiently reliable for safety-critical applications because (a)
infotainment. single-hop messages are not acknowledged and (b) a single-
In addition to the issue of proper modeling and specific hop message can be destroyed in a ‘collision’. In particular,
subsystem implementation, there needs to be a complete single-hop does not work in heavy traffic, where an explosion
system implementation. Besides those mentioned in Section of messages can easily overload the system.

325
 To make real progress, strong government leadership is re- [11] C. Maihöfer, T. Leinmüller, and E. Schoch, “Abiding geocast: time–
quired, although the manufacturers do not seem to want it. [12] stable geocast for ad hoc networks,” in Proceedings of the 2nd ACM
international workshop on Vehicular ad hoc networks, ser. VANET ’05.
In principle at least, high market penetration is possible with New York, NY, USA: ACM, 2005, pp. 20–29. [Online]. Available:
stand-alone units installed under mandatory vehicle manufac- http://doi.acm.org/10.1145/1080754.1080758
turing regulations. Initially, like safety belts and airbags, they [12] K. Matheus, R. Morich, and A. L§bke, “Economic background
of car-to-car communications,” in Informationssysteme f§r mobile
would be expensive. But with political will, a cooperative ITS Anwendungen (IMA 2004), October 2004. [Online]. Available: http:
system could reduce road casualties drastically by wrapping //www.network-on-wheels.de/downloads/IMA2004_Matheus_Paper.pdf
each vehicle in a very thick security blanket. Collisions would [13] B. Kerner, S. Rehborn, and H. Palmer, “Using probe vehicle data to
generate jam warning messages,” Traffic Engineering & Control, vol. 45,
become impossible, because cars would behave like bees in a no. 3, pp. 141 – 148, 2011.
swarm, adjusting autonomously to each others’ behavior. [14] M. Torrent-Moreno, “Inter-vehicle communications: Assessing
To sum up, almost all safety applications proposed so far information dissemination under safety constraints,” in 4th
Annual IEEE/IFIP Conference on Wireless On Demand Network
refer to a narrowly-defined hazard scenarios. There are dozens Systems and Services (WONS), Obergurgl, Austria, January
of them. In each case, output is generated in the same way, 2007. [Online]. Available: www.network-on-wheels.de/downloads/
by warning the driver of a threat. If only a small proportion TorrentMoreno_WONS07_AssessingInfoDissem.pdf
[15] Cable, “Free 4g and a freed up parking spot in nyc,” Cable, vol. 38,
of these applications ever made it onto the market, drivers no. 5, pp. 26 – 27, 2011. [Online]. Available: http://cable.poly.edu/
could be inundated with warnings. On the other hand, the most sites/default/files/download/cable_winter_2011_web.pdf
ambitious applications rely on 95-100% market penetration, [16] B. Song, D. Kim, and H. Choi, “Cooperative lateral control for automatic
which will require considerable political will to bring about, valet parking,” in Control, Automation and Systems (ICCAS), 2011 11th
International Conference on, oct. 2011, pp. 567 –570.
yet manufacturers don’t want government intervention. Once [17] C. Loper, A. Catala-Prat, J. Gacnik, J. Schomerus, and F. Koster,
installed, by far the biggest threats to a VANET system are “Vehicle automation in cooperation with v2i and nomadic devices
the security and reliability issues. communication,” in Intelligent Transportation Systems (ITSC), 2011
14th International IEEE Conference on, oct. 2011, pp. 650 –655.
[18] G. Le Lann, “Cohorts and groups for safe and efficient autonomous
R EFERENCES driving on highways,” in Vehicular Networking Conference (VNC), 2011
IEEE, nov. 2011, pp. 1 –8.
[1] T. B. Mello. (2011, September) Collision-avoidance systems can [19] H. Fubler, M. Torrent-moreno, M. Kruger, M. T. Rol, H. Hartenstein,
save lives. Bankrate.com. [Online]. Available: http://www.bankrate. and W. Effelsberg, “Studying vehicle movements on highways and their
com/finance/auto/collision-avoidance-systems-can-save-lives.aspx impact on ad-hoc connectivity,” ACM SIGMOBILE Mobile Computing
[2] D. Sedgwick. (2011, July) Study links low-speed collision-avoidance and Communications Review, Tech. Rep., 2005.
system to fewer accidents. [Online]. Available: http://www.autoweek. [20] P. Papadimitratos, G. Callandriello, J.-P. Hubaux, and A. Lioy, “Impact
com/article/20110719/CARNEWS/110719886 of vehicular communication security on transportation safety,” in IEEE
[3] R. Eigner and G. Lutz, “Collision avoidance in vanets - an application for MOVE. IEEE Computer Society, 2008.
ontological context models,” in Proceedings of the 5th IEEE Workshop [21] R. Baldessari, A. Festag, and J. Abeille, “Nemo meets vanet: A deploy-
on Context Modeling and Reasoning, Hong Kong, China, March 2008. ability analysis of network mobility in vehicular communication,” in
[4] A. Festag, G. Noecker, M. Strassberger, A. L§bke, B. Bochow, 7th International Conference on ITS Telecommunications (ITST 2007),
M. Torrent-Moreno, S. Schnaufer, R. Eigner, C. Catrinescu, and J. Ku- 2007, p. pages 375Ð380.
nisch, “Õnow Ð network on wheelsÕ: Project objectives, technology [22] G. Acosta-Marum and M. Ingram, “Six time- and frequency- selective
and achievements,” in Proceedings of 5rd International Workshop on empirical channel models for vehicular wireless lans,” Vehicular Tech-
Intelligent Transportation (WIT), Hamburg, Germany, March 2008, pp. nology Magazine, IEEE, vol. 2, no. 4, pp. 4 –11, dec. 2007.
pp. 211–216. [23] V. Milanes, J. Villagra, J. Godoy, J. Simo, J. Perez, and E. Onieva,
[5] E. Schoch, F. Kargl, T. Leinm§ller, and M. Weber, “Communication “An intelligent v2i-based traffic management system,” Intelligent Trans-
patterns in vanets,” IEEE Commun. Mag., vol. 46, pp. 119 – 125, portation Systems, IEEE Transactions on, vol. PP, no. 99, pp. 1 –10,
November 2008. 2012.
[6] M. Torrent-Moreno, A. Festag, and H. Hartenstein, “Sys- [24] S. Fujii, A. Fujita, T. Umedu, S. Kaneda, H. Yamaguchi, T. Higashino,
tem design for information dissemination in vanets,” in and M. Takai, “Cooperative vehicle positioning via v2v communications
Proceedings of 3rd International Workshop on Intelligent and onboard sensors,” in Vehicular Technology Conference (VTC Fall),
Transportation(WIT), Hamburg, Germany, March 2006, pp. 27 – 2011 IEEE, sept. 2011, pp. 1 –5.
33. [Online]. Available: http://www.network-on-wheels.de/downloads/ [25] D. Matolak, Q. Wu, J. Sanchez-Sanchez, D. Morales-Jimenez, and
WIT2006_InfoDissem_Torrent-Moreno_etal_wConfInfo.pdf M. Aguayo-Torres, “Performance of lte in vehicle-to-vehicle channels,”
[7] F. Dotzer, M. Strassberger, and T. Kosch, “Classification for traffic in Vehicular Technology Conference (VTC Fall), 2011 IEEE, sept. 2011,
related inter-vehicle messaging,” in Proceedings 5th IEEE International pp. 1 –4.
Conference on ITS Telecommuncations, June 2005. [26] M. K. Ajay Dureja, Aman Dureja, “Ieee 802.11 based mac improve-
[8] F. G. N. T. The CAMP Vehicle Safety Communications Consortium ments for manet,” IJCA Special Issue on MANETs, no. 2, pp. 54–57,
consisting of BMW, DaimlerChrysler and VW., “Vehicle safety com- 2010, published by Foundation of Computer Science.
munications project task 3 final report identify intelligent vehicle safety [27] H. Gossain, N. N, K. An, and D. P. Agrawal, “Supporting mac layer
applications enabled by dsrc,” U.S. Department of Transportation Na- multicast in ieee 802.11 based manets: Issues and solutions,” in In 29th
tional Highway Traffic and Safety Administration, 400 Seventh Street, Annual IEEE International Conference on Local Computer Networks
S.W. Washington, D.C. 20590, Tech. Rep. DOT HS 809 859, March (LCNÕ04, 2004, pp. 172–179.
2005. [28] S.-Y. Ni, Y.-C. Tseng, Y.-S. Chen, and J.-P. Sheu, “The broadcast
[9] A. K. K. Aboobaker and S. Wolthusen. (2010) Analysis of storm problem in a mobile ad hoc network,” in Proceedings of the 5th
authentication protocols in vehicular networks. [Online]. Avail- annual ACM/IEEE international conference on Mobile computing and
able: http://media.techtarget.com/searchSecurityUK/downloads/RHUL_ networking, ser. MobiCom ’99. New York, NY, USA: ACM, 1999,
Aboobaker_2010.pdf pp. 151–162. [Online]. Available: http://doi.acm.org/10.1145/313451.
[10] R. Eberhardt and C. Maihofer, “Virtual warning signs: a geocast enabled 313525
service for ad hoc networks,” in Proceedings of the 3rd Workshop [29] G. Korkmaz, E. Ekici, F. Özgüner, and U. Özgüner, “Urban multi-
on Applications and Services in Wireless Networks (ASWN), Bern hop broadcast protocol for inter-vehicle communication systems,” in
Switzerland, 2003, pp. 135 – 147. Proceedings of the 1st ACM international workshop on Vehicular ad hoc

326
 networks, ser. VANET ’04. New York, NY, USA: ACM, 2004, pp. 76– [50] M. Kihl, K. Bur, F. Tufvesson, and J. Aparicio Ojea, “Simulation
85. [Online]. Available: http://doi.acm.org/10.1145/1023875.1023887 modelling and analysis of a realistic radio channel model for v2v
[30] M. Sharma and G. Singh, “Evaluation of proactive, reactive and hybrid communications,” in Ultra Modern Telecommunications and Control
ad hoc routing protocol for ieee 802.11 mac and 802.11 dcf in vanet Systems and Workshops (ICUMT), 2010 International Congress on, oct.
using qualnet,” in CCSEA, e. a. D.C. Wyld, Ed. CS & IT-CSCP, 2011, 2010, pp. 981 –988.
p. 209Ð220. [51] A. Chelli and M. Patzold, “A non-stationary mimo vehicle-to-vehicle
[31] H. Fubler, M. Torrent-moreno, M. Transier, A. Festag, and H. Harten- channel model derived from the geometrical street model,” in Vehicular
stein, “Thoughts on a protocol architecture for vehicular ad-hoc net- Technology Conference (VTC Fall), 2011 IEEE, sept. 2011, pp. 1 –6.
works,” in 2nd Int. Workshop on Intelligent Transportation (WIT 2005), [52] T. Mangel and H. Hartenstein, “5.9ghz ieee 802.11p inter-vehicle com-
March 2005. munication: Non-line-of-sight reception under competition,” in Vehicular
[32] M. Kamal Abbas, M. Karsiti, M. Napiah, and B. Samir, “Traffic light Networking Conference (VNC), 2011 IEEE, nov. 2011, pp. 155 –162.
control via vanet system architecture,” in Wireless Technology and [53] B. M. Mughal, A. A. Wagan, and H. Hasbullah, “Analyzing safety bea-
Applications (ISWTA), 2011 IEEE Symposium on, sept. 2011, pp. 174 con loss rate in vanets with two-ray ground and nakagami propagation
–179. models,” in National Postgraduate Conference (NPC), 2011, sept. 2011,
[33] J. Lee and B. Park, “Development and evaluation of a cooperative pp. 1 –6.
vehicle intersection control algorithm under the connected vehicles [54] G. Dedes, D. Grejner-Brzezinska, D. Guenther, G. Heydinger,
environment,” Intelligent Transportation Systems, IEEE Transactions on, K. Mouskos, B. Park, and C. Toth, “Integrated gnss/inu, vehicle dy-
vol. PP, no. 99, pp. 1 –10, 2012. namics, and microscopic traffic flow simulator for automotive safety,”
[34] H. Rakha and R. Kamalanathsharma, “Eco-driving at signalized intersec- in Intelligent Transportation Systems (ITSC), 2011 14th International
tions using v2i communication,” in Intelligent Transportation Systems IEEE Conference on, oct. 2011, pp. 840 –845.
(ITSC), 2011 14th International IEEE Conference on, oct. 2011, pp. 341 [55] O. Shih, H.-M. Tsai, H.-M. Lin, and A.-C. Pang, “A rule-based mixed
–346. mobility model for cars and scooters (poster),” in Vehicular Networking
[35] (2010, June) Pilots in polish air disaster ignored warnings, transcripts Conference (VNC), 2011 IEEE, nov. 2011, pp. 198 –205.
reveal. Associated Press. [Online]. Available: http://www.guardian.co. [56] Y. Fang, F. Chu, S. Mammar, and M. Zhou, “Optimal lane reservation
uk/world/2010/jun/01/pilots-polish-ignored-warnings in transportation network,” Intelligent Transportation Systems, IEEE
[36] S. Schnaufer, H. Fuğler, M. Transier, and W. Effelsberg, “Vehicular ad- Transactions on, vol. PP, no. 99, pp. 1 –10, 2011.
hoc networks: Single-hop broadcast is not enough,” in Proceedings of [57] M. Di Felice, K. R. Chowdhury, and L. Bononi, “Cooperative spectrum
3rd International Workshop on Intelligent Transportatio (WIT), Ham- management in cognitive vehicular ad hoc networks,” in Vehicular
burg, Germany, March 2003, pp. 49 – 54. Networking Conference (VNC), 2011 IEEE, nov. 2011, pp. 47 –54.
[37] H. F§ğler, S. Schnaufer, M. Transier, and W. Effelsberg, “Vehicular ad- [58] K. Inage, S. Lee, T. Fujii, and O. Altintas, “White space vectors
hoc networks: From vision to reality and back,” in 4th Annual IEEE/IFIP for channel selection in vehicular cognitive networks,” in Vehicular
Conference on Wireless On Demand Network Systems and Services Networking Conference (VNC), 2011 IEEE, nov. 2011, pp. 55 –61.
(WONS), Obergurgl, Austria, January 2007. [59] E. del Arco, E. Morgado, J. Ramiro-Bargueno, I. Mora-Jimenez, and
[38] Y.-C. Chu and N.-F. Huang, “An efficient traffic information forwarding A. Caamano, “Vehicular sensor networks in congested traffic: Linking
solution for vehicle safety communications on highways,” Intelligent stv field reconstruction and communications channel,” in Intelligent
Transportation Systems, IEEE Transactions on, vol. PP, no. 99, pp. 1 Transportation Systems (ITSC), 2011 14th International IEEE Confer-
–13, 2011. ence on, oct. 2011, pp. 606 –613.
[39] S.-Y. Pyun, D.-H. Cho, and J.-W. Son, “Downlink resource allocation [60] V. Milanes, J. Godoy, J. Villagra, and J. Perez, “Automated on-ramp
scheme for smart antenna based v2v2i communication system,” in merging system for congested traffic situations,” Intelligent Transporta-
Vehicular Technology Conference (VTC Fall), 2011 IEEE, sept. 2011, tion Systems, IEEE Transactions on, vol. 12, no. 2, pp. 500 –508, june
pp. 1 –6. 2011.
[40] S. Abd El-atty, “Vehicular communications framework for efficient [61] M. Fogue, P. Garrido, F. Martinez, J.-C. Cano, C. Calafate, P. Manzoni,
multihop connectivity in ahvn,” in Vehicular Technology Conference and M. Sanchez, “Prototyping an automatic notification scheme for
(VTC Fall), 2011 IEEE, sept. 2011, pp. 1 –5. traffic accidents in vehicular networks,” in Wireless Days (WD), 2011
[41] C. Song, M. Liu, Y. Wen, J. Cao, and G. Chen, “Buffer and switch: IFIP, oct. 2011, pp. 1 –5.
An efficient road-to-road routing scheme for vanets,” in Mobile Ad-hoc [62] T. Menard, J. Miller, M. Nowak, and D. Norris, “Comparing the gps
and Sensor Networks (MSN), 2011 Seventh International Conference on, capabilities of the samsung galaxy s, motorola droid x, and the apple
dec. 2011, pp. 310 –317. iphone for vehicle tracking using freesim_mobile,” in Intelligent Trans-
[42] T. Alpcan and S. Buchegger, “Security games for vehicular networks,” portation Systems (ITSC), 2011 14th International IEEE Conference on,
Mobile Computing, IEEE Transactions on, vol. 10, no. 2, pp. 280 –290, oct. 2011, pp. 985 –990.
feb. 2011. [63] R. Juskalian. (2012, jan.) Europe’s driverless car (driver still required).
[43] W.-C. Shieh, S.-I. Sou, and S.-Y. Tsai, “A study of video frame sharing Online. [Online]. Available: http://m.technologyreview.com/business/
in sparse vehicular networks,” in Parallel and Distributed Systems 39410/
(ICPADS), 2011 IEEE 17th International Conference on, dec. 2011, pp.
444 –448.
[44] D. Hambling. (2011, March) Gps chaos: How a $30 box can
jam your life. [Online]. Available: http://www.newscientist.com/article/
dn20202-gps-chaos-how-a-30-box-can-jam-your-life.html
[45] I. A. Sumra, H. Hasbullah, and J.-l. A. Manan, “Vanet security research
and development ecosystem,” in National Postgraduate Conference
(NPC), 2011, sept. 2011, pp. 1 –4.
[46] R. Lu, X. Lin, X. Liang, and X. Shen, “Flip: An efficient privacy-
preserving protocol for finding like-minded vehicles on the road,”
in Global Telecommunications Conference (GLOBECOM 2010), 2010
IEEE, dec. 2010, pp. 1 –5.
[47] R. J. Hwang, Y.-K. Hsiao, and Y.-F. Liu, “Secure communication scheme
of vanet with privacy preserving,” in Parallel and Distributed Systems
(ICPADS), 2011 IEEE 17th International Conference on, dec. 2011, pp.
654 –659.
[48] J. Petit and Z. Mammeri, “Analysis of authentication overhead in
vehicular networks,” in Wireless and Mobile Networking Conference
(WMNC), 2010 Third Joint IFIP, oct. 2010, pp. 1 –6.
[49] R. A. Uzcategui and G. Acosta-Marum, “Wave: A tutorial,” IEEE
Communications Magazine, vol. 47, no. 5, pp. 126–133, May 2009.

327