Ad Hoc Networks 8 (2010) 457–461

Contents lists available at ScienceDirect

Ad Hoc Networks
journal homepage: www.elsevier.com/locate/adhoc

Guest Editorial

Vehicular networks

The advent of ad hoc wireless networking is arguably ing the specific requirements for developing different
one of the most significant developments in wireless net- VANET applications. We note here that there are several
working and telecommunications in the last decade. While excellent surveys and tutorials on Vehicular Ad Hoc Net-
the research into this area has started as a result of the works some of which have influenced our viewpoint (see,
immediate needs of the Department of Defense (DoD) in e.g., [2]). The research in this area is relatively new and it
the USA for military combat operations in hostile territo- essentially spans the last decade (2000-2010). While some
ries, the application areas have since grown tremendously significant progress has been made, it is fair to say that
and have expanded to include the synergistic area of sen- several key research and development problems remain
sor networks as well. Nevertheless, it is fair to say that open.
the bulk of the research in ad hoc networks has stayed fo- In terms of vehicle-to-vehicle (V2V) communications,
cused on military applications with few commercial appli- the use of both radio (very high frequency (i.e., VHF), mi-
cations that seemed viable in the foreseeable future. This cro, and millimeter waves) and infrared waves have been
picture has changed dramatically in the last 5 years or so reported in experimental V2V systems. While infrared
with the advent of the automobile industry showing inter- and millimeter waves can support only line-of-sight com-
est in the future of Vehicular Ad Hoc Networks, mainly for munications, VHF and microwaves can support broadcast
safety applications. Indeed, this is probably the biggest communications as well. VHF can provide long links but
new commercial application of ad hoc networks with real at low speed and for this reason the mainstream mode of
and concrete applications (such as safety) driving the communications is to use microwaves. In the United
march of the underlying technology. The fact that FCC States, 75 MHz in the 5.9 GHz band is allocated for
has allocated spectrum in the 5.9 GHz region to this new VANETs; in Europe and Japan, the spectrum allocated to
application clearly indicates that this is much more than VANETs is in the 5.8 GHz band. In Europe, 10 MHz also is
a fad and it is destined to grow into a commercially viable available between 2010 and 2020 MHz for vehicle
technology. communications.
When one reflects upon the current state of affairs, it is In the first paper of this Special Issue, Tacconi, Miorandi,
probably not very surprising that we are witnessing an Chiti, and Fantacci propose a system architecture for en-
exponentially growing interest in this area, both from aca- abling mobile nodes to query a largely deployed sensor
demia and automobile industry: after all, with 60–100 network in an intelligent transportation system scenario.
embedded sensors with their corresponding microproces- The authors first identify three different types of nodes
sors, the contemporary cars are the best mobile computing in the network: mobile sinks (i.e., the nodes moving and
platforms that one could dream for. They are highly mobile querying the WSN), vice-sinks (i.e., nodes able to commu-
and they have a tremendous amount of embedded com- nicate directly with mobile sinks) and ordinary sensor
puting power. While the car industry, for safety applica- nodes (i.e., nodes sensing a phenomenon and communicat-
tions, has embraced this new research area and ing in a multi-hop fashion). Then, they present protocols
technology as its own, at the present juncture the number and algorithms specifically tailored to such a scenario, in
of potential applications have quickly expanded beyond particular at the MAC and network layers. The proposed
safety and now includes other types of applications as architecture lends itself nicely to situations in which WSNs
well. deployed in a parking place or along a road, providing to
Vehicular Ad Hoc Networks (VANETs) can be consid- cars information on the conditions of the surrounding
ered as a subset or special case of Mobile Ad Hoc Networks environment. In addition, the authors introduce and eval-
(MANETs) which have been studied extensively in the lit- uate a simple geographic routing protocol and two differ-
erature (see, for example, [1] and the references therein). ent load balancing techniques. Results show that the
In this editorial, we present an overview of the current VA- proposed solutions enable the introduction of novel intel-
NET research at all layers of the protocol stack, emphasiz- ligent transportation system applications.

1570-8705/$ - see front matter Ó 2010 Published by Elsevier B.V.

doi:10.1016/j.adhoc.2010.01.001
458 Guest Editorial / Ad Hoc Networks 8 (2010) 457–461

Most current work on V2V and vehicle-to-infrastruc- access networks include quick adaptability to frequent
ture (V2I) systems employ the 802.11a, b, or g technology. changes in the network topology due to vehicular mobility
The physical layer (PHY) of the IEEE 802.11p standard is and delay awareness in data delivery. To address these
based on orthogonal frequency-division multiplexing challenges, the authors propose a cross-layer position-
(OFDM). This physical layer will be very similar to the based delay-aware communication protocol called
physical layer of 802.11a. The major difference will be PROMPT. This protocol adopts a source routing mechanism
that the bandwidth is halved in order to decrease the in- that relies on positions independent of vehicle movement
ter-symbol interference caused by the multipath delay rather than on specific vehicle addresses. Vehicles monitor
spread and the Doppler spread effect. The 802.11p targets information exchange in their reception range to obtain
a transmission range between 300 m and 1 km. The data data flow statistics, which are then used in estimating
throughput ranges from 3 to 27 Mbits/s. At the PHY/ the delay and selecting best available paths. Via extensive
MAC level, random access techniques seem prevalent simulations the authors show that PROMPT outperforms
and, more precisely, the IEEE 802.11 Distributed Coordi- the existing routing protocols for VANETs in terms of
nation Function (DCF) scheme is used. While random ac- end-to-end packet delay, packet loss rate, and fairness of
cess protocols and DCF scheme currently seem to be the service.
preferred choices for MAC protocols in VANETs, in gen- The paper by Sommer, Schmidt, Chen, German, Koch,
eral, MAC design seems to be an area which requires fur- and Dressler addresses the difficulties VANETs might have
ther research. The key factors which complicate the initially at the introduction stage of the DSRC technology,
design of good MAC protocols include high variability in namely the low penetration rate and its repercussions.
vehicle density, sharing time and space simultaneously, The authors argue that while intelligent transportation
etc. Such considerations might ultimately necessitate systems (ITS) are a hot topic in the communications soci-
adaptive MAC protocols for VANETs. While the current ety, currently, research is primarily focusing on setting
Dedicated Short Range Communications (DSRC) standard up VANETs based on WLAN technology. They also claim
has specified the PHY and MAC Layers of V2V communi- that, VANETs are heavily dependent on market penetration
cations, the aforementioned issues might lead to radically or infrastructure support and, therefore, Third-Generation
different MAC protocols for VANETs. The jury is out on (3G) networks might complement these efforts as they
this issue as well. are already widely deployed and can serve as the basis
At the network layer, much work remains to be done. In for Car-to-Infrastructure (C2I) applications. Sommer et al.
terms of routing, for example, while many people consider develop a simulation framework for holistic analysis of
geographic routing as the right approach for routing in VA- complex UMTS-based ITS. This framework couples simula-
NETs, hitherto this has not been established in a compre- tion models with corresponding protocols of the UMTS link
hensive manner for safety and non-safety applications. level, of higher network layers, and of road traffic. Based on
One of the obstacles is the lack of detailed studies on the their simulation framework, the authors evaluate a UMTS-
impact of mobility of mobility on propagation in urban based traffic information system (TIS) in a typical highway
and highway scenarios. In the absence of such in-depth scenario in which information about traffic jams needed to
studies, to design good VANET protocols is quite difficult. be communicated to other cars for optimized route plan-
While some progress has been made in this area over the ning. The authors show the availability of UMTS multicast
last few years (see, e.g., [3]), it is fair to say that, compared distribution services are demanded for an efficient opera-
to well-known ad hoc routing protocols (such as DSR and tion of the TIS application.
AODV) used in MANET, a widely accepted VANET routing Another important area that needs major attention of the
protocol does not exist yet. research community is data dissemination in VANETs. It is
The paper by Lee, Cheng, and Gerla propose GeoCross, clear that the underlying techniques that need to be devel-
a simple, yet novel, event-driven geographic routing pro- oped for data dissemination in VANETs will be application
tocol that removes cross links dynamically to avoid rout- driven. This implies that data dissemination techniques
ing loops in urban Vehicular Ad Hoc Networks (VANETs). for safety applications and non-safety applications could
GeoCross exploits the natural planar feature of urban potentially be very different. While the car industry has
maps without resorting to cumbersome planarization. been looking into how to do data dissemination for safety
Its feature of dynamic loop detection makes GeoCross application over the last decade, it is quite interesting to
suitable for highly mobile VANET. The authors show that see that recently companies like Google also started getting
in pathological cases, GeoCross’s packet delivery ratio involved in data dissemination with the unveiling of their
(PDR) is consistently higher than Greedy Perimeter State- smart phone technology Nexus One which is running on An-
less Routing’s (GPSR’s) and Greedy Perimeter Coordinator droid Operating System (OS). In the latter case, it is clear that
Routing’s (GPCR’s). They also show that caching (Geo- companies like Google are interested in the lucrative busi-
Cross + Cache) provides the same high PDR but uses fewer ness opportunity attached to mobile advertising and en-
hops. abling a platform for third party application developers.
The second paper on routing issues in the Special Issue The underlying key challenges for developing efficient data
is the paper by Jarupan and Ekici. In this paper, the authors dissemination techniques remain to be the high mobility
focus on the design of communication protocols for vehic- and the intermittent connectivity of VANETs. Another open
ular access networks where vehicles access a wired back- research issue is whether these challenges should be ad-
bone network by means of a multi-hop data delivery dressed at the PHY layer, MAC layer, or application layer
service. Key challenges in designing protocols for vehicular for efficient data dissemination.
 Guest Editorial / Ad Hoc Networks 8 (2010) 457–461 459

The paper by Lochert, Scheuremann, and Mauve be met (such as the real world connectivity a VANET can
propose an algorithm for the hierarchical aggregation of provide as well as the connection duration between vehi-
observations in dissemination-based, distributed traffic cles on US roads) in order to enable multiplayer games
information systems. Instead of transmitting observed over VANETs. The authors propose a new paradigm of mul-
parameters directly, the authors propose soft-state tiplayer games over VANETs, one which utilizes the new,
sketches—an extension of Flajolet-Martin sketches—as a interactive and dynamic characteristics of the VANET envi-
probabilistic approximation. This data representation is ronment, while adapting to its inherent constraints. The
duplication insensitive, a trait that overcomes two central new business opportunity and the challenges for this
problems of existing aggregation schemes for VANET new pervasive computing and communication application
applications. First, when multiple aggregates of observa- are also highlighted in the paper.
tions for the same area are available, it is possible to com- In addition to the need for a new protocol stack (or
bine them into an aggregate containing all information modified TCP/IP stack) for VANETs, another issue which
from the original aggregates. This is fundamentally differ- has not been addressed by the research community so
ent from existing approaches where typically one of the far is how to approach VANET problems in a holistic man-
aggregates is selected for further use while the rest is dis- ner with a general architecture. Such a holistic approach
carded. Second, any observation or aggregate can be in- and general architecture may require new conceptual
cluded into higher level aggregates, regardless if it has and practical modeling tools that can account for the un-
already been previously—directly or indirectly—added. ique mobility and connectivity characteristics of VANETs
The authors show through extensive simulations that (see, e.g., [4]). Unfortunately, most of the reported research
these characteristics result in a very flexible aggregate con- to date takes a piecemeal approach, focusing on the spe-
struction and a high-quality of aggregates. cific problem investigated. The absence of a general VANET
Another important open research issue concerns identi- architecture and a suitable protocol stack is not conducive
fying the right protocol stack for VANETs. While it is clear to cross-layer optimization approaches which can lead to
that TCP/IP stack is not well-suited to VANETs, the alterna- better solutions. Hence, much work remains to be done
tive does not seem clear. Should the research community in this domain as well.
look for a new protocol stack that is suitable for VANETs Today, there are several uncertainties in this exciting
or should it try to adapt TCP/IP to VANET domain? Ques- new research area that are unveiling as we go along. One
tions like these are not easy to answer and require major of these pertains to the fact that, so far the VANET research
new research efforts. Given the widely accepted status of for safety applications is using a paradigm which is known
TCP/IP protocol, it might be hard to justify a new protocol as zero-infrastructure networking. One could argue that
stack unless, of course, substantial advantages and benefits there are good reasons for this as the wireless service pro-
can be demonstrated. viders do not see the business potential for getting in-
Similar to security concerns in the MANETs domain, volved. This situation, however, might soon change as
security issues and challenges for VANETs seem important, the use of VANET technology seems to be a very serious
non-trivial, and urgent. The specific security issues such as alternative even for traffic information systems (TIS). If
privacy, authentication, integrity, and non-repudiation are one combines this with other user applications that are
serious issues that call for serious and practical solutions. coming to the forefront (such as Internet Access and enter-
The proposed security solutions have to take into account tainment), it is conceivable to think that wireless service
the degradation such solutions might cause in key perfor- providers may soon also get involved and start looking at
mance metrics such as end-to-end delay, throughput, this as a lucrative new business opportunity. This in turn
bandwidth, etc. in addition to cost and complexity. might accelerate the adoption of the new DSRC technology
Although some progress has been made in identifying by more car manufacturers thus enabling a growing num-
the key security challenges and proposing solutions for ber of new applications which rely on the ad hoc co-oper-
them, much work remains to be done in this area to ensure ation of cars on a highway or in an urban scenario.
that the rather complex security requirements of VANETs Researchers in academia and industry are currently look-
can be satisfied. ing into different ways and alternatives for expediting this
While several interesting VANET applications are cur- process.
rently being pursued, generically speaking, there are two We would like to express our deep appreciation to Dr.
major categories of applications; namely, safety and user Ian Akyildiz, the Editor-in-Chief of Ad Hoc Networks Jour-
applications. The main focus of safety applications is to nal, for giving us the opportunity to put together this Spe-
use VANETs for increasing the safety of drivers and to pre- cial Issue on Vehicular networks. Without his vision and
vent or mitigate traffic congestion when accidents happen. generous help this issue would not have been possible.
User applications, on the other hand, refer to value-added We would like to thank all the researchers who have sub-
services such as entertainment and/or Internet access mitted their contributions to this Special Issue. We are also
(connectivity). very grateful to several reviewers in different countries
The paper by Tonguz and Boban investigate the possi- (and continents) who took it upon themselves to provide
bility of a new type of application, namely multiplayer detailed and constructive reviews for ensuring the high-
games, in a Vehicular Ad Hoc Network (VANET) Environ- quality of the set of papers accepted. It is our hope that
ment. First, the authors analyze the available empirical the telecommunications and networking community will
data on travel and traffic volume in the United States, enjoy reading this Special Issue and use it as a valuable ref-
and point out the most important challenges that have to erence for their future work.
460 Guest Editorial / Ad Hoc Networks 8 (2010) 457–461

References Technical Program Committees of several IEEE conferences

(such as INFOCOM, SECON, GLOBECOM, ICC, VTC, WCNC)
[1] O.K. Tonguz, G. Ferrari, Ad Hoc Wireless Networks: A and symposia in the area of wireless communications and
Communication-Theoretic Perspective, Wiley, 2006. optical networks, Dr. Tonguz currently serves or has served
[2] Y. Toor, P. Muhlethaler, A. Laouti, A. De La Fortelle, Vehicle ad hoc
networks: applications and related technical issues, IEEE as an Associate Editor for the IEEE Transactions on Com-
Communications Surveys and Tutorials 10 (3) (2008) 74–88. munications, IEEE Communications Magazine, IEEE Journal
[3] O.K. Tonguz, N. Wisitpongphan, F. Bai, DV-CAST: a distributed of Lightwave Technology, and Ad Hoc Networks Journal
vehicular broadcast protocol for vehicular ad hoc networks, IEEE (Elsevier). He was a Guest Editor of the Special Issue of the
Wireless Communications Magazine, (2010), in press.
[4] O.K. Tonguz, W. Viriyasitavat, F. Bai, Modeling urban traffic: a cellular IEEE Journal of Lightwave Technology and IEEE Journal on
automata approach, IEEE Communications Magazine 47 (5) (2009) Selected Areas in Communications on Multiwavelength
142–150. Optical Networks and Technology, published in 1996, and a
Guest Editor of the Special Issue of JOURNAL of Mobile
Multimedia on Advanced Mobile Technologies for Health
Ozan K. Tonguz received the B.Sc. Care Applications (2006).
degree from the University of Essex,
England, and the M.Sc. and the Ph.D.
degrees from Rutgers University, NJ, Fan Bai is a Senior Researcher in the
respectively, all in electrical engi- Electrical & Control Integration Lab.,
neering. He currently serves as a Research & Development and Plan-
tenured Full Professor in the Depart- ning, General Motors Corporation,
ment of Electrical and Computer since Sep., 2005. Before joining Gen-
Engineering at Carnegie Mellon Uni- eral Motors research lab, he received
versity (CMU). Before joining CMU in the B.S. degree in automation engi-
August 2000, he was with the ECE neering from Tsinghua University,
Dept. of the State University of New York at Buffalo (SUNY/ Beijing, China, in 1999, and the
Buffalo). He joined SUNY/Buffalo in 1990 as an Assistant M.S.E.E. and Ph.D. degrees in electri-
Professor, where he was granted early tenure and promoted cal engineering, from University of
to Associate Professor in 1995, and to Full Professor in 1998. Southern California, Los Angeles, in 2005. His current
Prior to joining academia, he was with Bell Communications research is focused on the discovery of fundamental prin-
Research (Bellcore) between 1988 and 1990 doing research ciples and the analysis and design of protocols/systems for
in optical networks and communication systems. His cur- next-generation Vehicular Ad hoc Networks (VANETs), for
rent research interests are in Vehicular Ad Hoc Networks safety, telematics and infotainment applications. Dr. Bai has
and sensor networks, high-speed networking (Internet), published about 40 book chapters, conference and journal
wireless networks and communication systems, optical papers, including INFOCOM, MobiHoc, SECON, ICC, Globe-
communications and networks, satellite communications, com, WCNC, JSAC, IEEE Wireless Communication Magazine,
bioinformatics, and security. He has published about 300 IEEE Communication Magazine and Elsevier AdHoc Net-
technical papers in IEEE journals and conference proceed- works Journal. In 2006, he received Charles L. McCuen
ings. He is well-known for his contributions in wireless Special Achievement Award from General Motors Corpora-
communications and networks as well as optical commu- tion ‘‘in recognition of extraordinary accomplishment in
nications and networks. His recent work on iCAR (the Inte- area of vehicle-to-vehicle communications for drive assis-
grated Cellular and Ad Hoc Relay Systems) is internationally tance & safety.” He serves as Technical Program Co-Chairs
acclaimed as well. He is the author (with G. Ferrari) of the for IEEE WiVec 2007 and IEEE MoVeNet 2008, and serves as
Wiley book entitled ‘‘Ad Hoc Wireless Networks: A Com- guest editors for IEEE Wireless Communication Magazine,
munication-Theoretic Perspective”. He was also the archi- IEEE Vehicular Technology Magazine and Elsevier AdHoc
tect of the ‘‘High Performance Waveform (HPW)” that was Networks Journal. He is also serving as a Ph.D. supervisory
implemented in Harris RF Communications’ AN/PRC-117f committee member at Carnegie Mellon University.
UHF band man-pack tactical radio. His industrial experience
includes periods with Bell Communications Research, CTI
Inc., Harris RF Communications, Aria Wireless Systems, Cem U. Saraydar received a bache-
Clearwire Technologies, Nokia Networks, Nokia Research lor’s degree from Bogazici Univer-
Center, Neuro Kinetics, Asea Brown Boveri (ABB), General sity, Istanbul, Turkey, and the M.S.
Motors (GM), and Intel. He currently serves or has served as and Ph.D. degrees from WINLAB,
a consultant or expert for several companies (such as Aria Rutgers University, all in electrical
Wireless Systems, Harris RF Communications, Clearwire engineering. Following his PhD, he
Technologies, Nokia Networks, Alcatel, Lucent Technolo- worked for the Performance Analysis
gies), major law firms (Jones Day, WilmerHale, Williams and Department at Bell Laboratories,
Connolly, Heller Ehrman, Baker Botts, etc.), and government Holmdel, as a Member of Technical
agencies in the USA, Europe, and Asia in the broad area of Staff and at the ECE Department at
telecommunications and networking. He is also a Co- NJIT, Newark, NJ as a Research
Director (Thrust Leader) of the Center for Wireless and Associate/Lecturer. Since 2005, he is with General Motors
Broadband Networking Research at Carnegie Mellon Uni- R&D, Warren, MI, where he is a Lab Group Manager for the
versity. More details about his research interests, research Wireless Systems and Technology group. His current
group, projects, and publications can be found at http:// research interests include wireless sensor networks and
www.ece.cmu.edu/~tonguz/. In addition to serving on the wireless ad hoc networking.
 Guest Editorial / Ad Hoc Networks 8 (2010) 457–461 461

Lead Guest Editor Guest Editor

Ozan K. Tonguz Cem U. Saraydar
Carnegie Mellon University, United States General Motors, United States
E-mail address: tonguz@ece.cmu.edu E-mail address: cem.saraydar@gm.com
Guest Editor
Available online 16 January 2010
Fan Bai
General Motors, United States
E-mail address: fan.bai@gm.com