Preliminary Results for Secure Traffic Regulation
Assia Belbachir
1,3
, Sorore Benabid
1,3
, Marcia Pasin
2
and Amal El Fallah Seghrouchni
3
1
Institut Polytechnique des Sciences Avanc
´
ees, IPSA, Ivry-sur-Seine, France
2
Universidade Federal de Santa Maria, Santa Maria, Brazil
3
Sorbonne Universit
´
e, Paris, France
Keywords:
Secure Traffic Regulation, SUMO, Central Communication Coordinator (CCC).
Abstract:
In this paper, we propose a solution for traffic regulation in Intelligent Transportation Systems (ITS). Due to
the heterogeneity and dynamic nature of transportation networks, the big challenge is to be able to take the
correct decisions to smooth the traffic flow. This decision is at the same time centralized within a Central
Communication Coordinator (CCC) which is in charge of controlling the traffic lights and also distributed
among the vehicles that are in the nearby of cross section. The decision should be made upon a permanent
exchange of information between the vehicles and the CCC’s. Thus, an important issue is to ensure a robust
and secure communication between the different components able to resist to hypothetic hacker’s attacks. The
aim of this paper is to setup an intelligent system to regulate traffic flow taking into account different types
of attacks. We show in this paper our simulated results using the client MATLAB together with the server
SUMO (Simulation of Urban Mobility) so that client can access and modify the simulation environment. We
simulated one intersection using an algorithm to regulate the traffic and explain our obtained results.
1 INTRODUCTION
Traffic congestion is a major constraint nowadays. In
many big cities, traffic jams are daily reported, and
chaotic transit in the main avenues during rush hours
is usual. The situation is even worse in the begin-
ning and the end of holidays. During such traffic jams,
which can exceed hundreds of kilometers, people lost
many hours in their journey just waiting in their cars.
There is a lot of cause for traffic jam, one of them
is caused by traffic lights (TLs). The latter can gener-
ates traffic deadlock in an intersection and can affect
other intersections. Therefore, TLs can be removed
to improve traffic
1
. Drivers waste time waiting at the
red light.
However, TLs can also be used to regulate traf-
fic and to reduce traffic jams at road intersections.
Recently TLs were installed to reduce traffic jams in
an intersection in Nantes, France
2
. Basically, in this
case, the TL regulates the traffic when a collector road
reaches the motorway. TLs were also introduced with
1
https://www.aol.co.uk/2017/02/14/paris-removes-
traffic-lights-to-fix-congestion-and-improve-safety/
2
https://www.ouest-france.fr/pays-de-la-loire/nantes-
44000/circulation-les-feux-anti-bouchon-s-allument-ce-
lundi-nantes-4917955
the same objective in Grenoble, France
3
. Moreover,
one third of the accidents in France are caused by the
fact that the drivers ignore the red signal.
In a near future, with the fully implementation of
the Vehicular Ad hoc NETworks (VANETs) technol-
ogy, vehicles will exchange messages and collaborate
to achieve more efficient strategies to deal with in-
tersection control and route planning, and to possi-
bly collaborate to reduce time travels and traffic jams.
VANETs pose new challenges to transportation net-
works (Eze and Liu, 2014). For instance, maintain-
ing a huge set of cars communicating with each other
is not a trivial task. Due to the heterogeneity of Ve-
hicle to Vehicle (V2V) and Vehicle to Infrastructure
(V2I) communication devices and the complexity of
the exchanged information, it is difficult to formal-
ize, to model and to design solutions to achieve the
required needs.
Therefore, the need for standardized solutions to
deal with this heterogeneity is an issue, since these
devices will provide information consumed by other
VANETs services.
On the other hand, several potential threats to ve-
3
https://france3-regions.francetvinfo.fr/auvergne-
rhone-alpes/feux-regulation-limiter-bouchons-rocade-sud-
grenoble-1082649.html
Belbachir, A., Benabid, S., Pasin, M. and Seghrouchni, A.
Preliminary Results for Secure Traffic Regulation.
DOI: 10.5220/0006904504190424
In Proceedings of the 15th International Conference on Informatics in Control, Automation and Robotics (ICINCO 2018) - Volume 2, pages 419-424
ISBN: 978-989-758-321-6
Copyright © 2018 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
419

hicular communication network exist, ranging from
fraudulent messages capable of disrupting traffic, or
even causing danger to driver or to its environment.
Traffic infrastructure and involved vehicles must be
resilient against intruders that potentially generate
disruption, degrade safety, or gain an unfair advan-
tage (Ghena et al., 2014).
In this direction, this paper proposes a novel solu-
tion for secure traffic regulation. We define the con-
cept of a Central Communication Coordinator (CCC),
a trust unite which is responsible to regulate the traf-
fic in one intersection (see Figure 1). A CCC can be
implemented in the infrastructure and will get infor-
mation from the cars and other infrastructures which
are in the vicinity of a given intersection (e.g. traffic
light). The problem here is that a CCC regulates the
traffic for only one intersection. Due to its decision it
can generate other traffic jams for other intersections.
Thus, the CCCs need to communicate with each other.
However, such exchange of information may cause
some overhead due the need of communication and
processing. Thus, CCCs need to decide whether to
communicate all other CCCs or with a restricted num-
ber of CCCs. In this paper, we are investigating only
one intersection. The focus is on the communication
protocol among cars and a given CCC.
CCC
communication
secure
Figure 1: The Central Communication Coordinator (CCC)
and vehicles with communication support in a one intersec-
tion scenario.
Additionally, cars should be able to organize
themselves and exchange information in a secure way.
In this sense, our paper defines a different secure
way to exchange information between cars, CCCs and
vice-versa. Then, we propose a solution to avoid
deadlock for traffic jam with regard to one intersec-
tion.
We evaluate our proposal through simulation. For
our simulations we used SUMO (Simulation of Ur-
ban Mobility) (Behrisch et al., 2011). SUMO is an
open source road traffic simulation package designed
to model the flow of vehicle traffic by treating differ-
ent realistic cases associated with MATLAB. SUMO
can handle large environments, with complex struc-
tures using 10.000 streets, and it can import many net-
work formats such as XML for example.
The paper is divided into five sections. Section
two, state the actual work on traffic regulation. The
third part explain the contribution of our work using
one intersection and a secure way of data exchange.
The fourth part explain our evaluation and simulated
results. Finally, we conclude our work.
2 STATE OF THE ART
Previous researches have suggested VANETs as a way
to enable V2V communications and to allow informa-
tion exchange and other types of information among
vehicles (Rawat et al., 2014) (Chembe et al., 2017).
Several communication protocols have been proposed
for V2V and V2I communication such as Wireless
Access in Vehicular Environments (WAVE) and Ded-
icated Short Range Communications (DSRC). WAVE
and DSRC standards are defined in IEEE 1609.1-4
and 802.11p respectively. The U.S. Federal Commu-
nication Commission (FCC) has allocated, in 1999,
dedicated 75 MHz frequency spectrum in the range
5.85 GHz to 5.925 GHz to be used exclusively for
V2V and V2I communication. The DSRC spectrum
is divided into seven channels with a 10 MHz band-
width allocated to each one. Six out of these channels
are service channels (SCH) and the center one is the
control channel (CCH).
However, security aspects need to be taken into
account for traffic regulation and VANETs. A sur-
vey by (Karagiannis et al., 2011) introduced the basic
characteristics of VANETs, and provided an overview
of applications and their requirements, along with
challenges and proposed solutions. The authors
highlight vehicle communication security capabili-
ties such as mechanisms to maintain privacy and
anonymity, integrity and confidentiality, resistance to
external attacks, authenticity of received data, data
and system integrity.
(Hasrouny et al., 2017) focus on VANET security
characteristics and challenges as well as existing so-
lutions. They present a classification of the different
attacks a VANETs may suffer such as authentication,
availability by resisting to DoS (Denial of Service) at-
ICINCO 2018 - 15th International Conference on Informatics in Control, Automation and Robotics
420

tacks, confidentiality, non-repudiation, integrity, pri-
vacy, data verification, access control, traceability and
revocability, error detection, liability identification,
flexibility and efficiency despite of timing constraints.
Moreover, VANETs need to be resilient against a se-
ries of attacks including: DoS, malwares, spam, man
in the middle attack, injection of erroneous messages,
cheating with position information, etc.
In fact, cars with selfish behavior and malicious at-
tacks may impact on the expected traffic performance.
For instance, a malicious node can steal frames trans-
mitted from cars or from the infrastructure. A mali-
cious node can also propagate a malicious message to
disturb traffic behavior. Selfish cars can take benefit
of cheating aiming to arrive first. Fortunately, due to
the special characteristics of VANETs and their dy-
namic nature, cheating cars cannot choose which cars
they will interact. Thus, spread lies might become
irrelevant as more participating cars receive recent in-
formation (Lin et al., 2007). However, it is not yet
clear how malicious nodes will influence in intersec-
tion control protocols.
(Lin and Li, 2013) proposed a cooperative authen-
tication scheme for VANETs using a reward approach
implemented by a Trusted Authority (TA). The TA re-
ceives the messages from vehicles when vehicles pass
by the Road Side Units (RSUs), and it sends back new
messages to the vehicles based on their past rewards.
The vehicles obtain a reward as they make contribu-
tion to the network.
A more recent work (Lim and Manivannan, 2016)
presents a protocol for fast dissemination of au-
thenticated messages to propagating phenomena in
VANETs. Phenomena messages are used to prop-
agate accidents, road conditions, etc. RSUs dis-
seminate authenticated messages about the observed
phenomena by vehicles in RSU transmission range.
RSUs are TAs and can verify the authenticity of the
sender and the message integrity before message dis-
semination. Messages sent by vehicles do not require
authentication and verification by other vehicles. The
aim of the protocol is to ensure the anonymity of the
senders and also to allow a mechanism to trace the
messages, when required, for law enforcement agen-
cies, for instance.
Centralized system represents single point of fail-
ures (SPOF) and, therefore, worthwhile targets for at-
tackers. Alternatively, decentralized solutions based
on new technologies can be proposed.
In this direction, blockchain is a promising dis-
tributed architecture to build reliable solutions to In-
telligent Transportation Systems (ITS). In Block-VN
(Sharma et al., 2017), participating entities have dif-
ferent behaviors. Controller nodes provide necessary
services on a large scale communication, miner nodes
deal with communication issues, and other nodes are
just ordinary nodes (vehicles, for instance). Ordinary
nodes send a service request message to other vehi-
cles or for the controller nodes. Using this communi-
cation hierarchy in a distributed way, scalability and
high availability are expected to be achieved. Block-
VN also aims to improve security using trusted inter-
mediary services and by providing distributed, secure,
and shared records of all system actions.
(Leiding et al., 2016) also suggests the application
of blockchain to deal with ITS challenges. Traffic
regulations algorithms can be implemented with the
support of Ethereum high-level languages. Based on
information about cars and traffic conditions, it is pos-
sible to identify and punish misbehaving cars. This
might include high speeding, ignoring traffic lights,
causing an accident, etc. Each car will be identified
by its unique public key. Thus, a punishment can be
imposed to the car (or driver) with such correspond-
ing Ethereum account.
However, the blockchain theory is not yet mature.
Scalability is still a criterion to be more investigated
further. Another important issue is that VANETs are
subject to intense churn (nodes are coming in and out
all the time) and it is not yet known if this behavior
will be efficiently handled by the blockchain technol-
ogy.
3 SECURE TRAFFIC
REGULATION STRATEGY
In this section, we explain the main contribution of
our work using one intersection and a secure way of
data exchange.
3.1 Traffic Regulation Algorithm
The CCC is a TA and communicates with vehicles and
the infrastructure. The used infrastructure is the TL.
The CCC controls the TLs state duration and transi-
tion according to the ratio (rap) between flows. Al-
gorithm 1 represents the used function to regulate the
TL for one intersection. The number of vehicles on
the entrance (E) and the exit (S) of a road for each di-
rection (North, South, East and West) is given in the
following variables: EN, ES, EE, EO, .. ., etc. rap is
computed using the ratio between the vehicle’s num-
ber which is entering to the intersection in a vertical
lane (EN, ES) and the horizontal lane (EE, EO) (see
Figure 3).
The counter shows how long the TL has not
changed its state. GreeenLight (NSALL) and Green-
Preliminary Results for Secure Traffic Regulation
421

Light (EOALL) represent respectively the green light
state for the two TLs in the vertical lane (EN, ES) and
the horizontal lane (EE, EO). We predefined a mini-
mum and a maximum TL state to green color as fol-
low: MIN and MAX u.t (unit of time). The procedure
is changing the traffic light color according to the ratio
(rap) and the counter (C) duration. When the traffic
is dense at EN and ES compared to EE and EO the
green light state is activated. Otherwise, and accord-
ing to rap value the green light state of EN and ES is
activated.
Algorithm 1: Traffic light regulation algorithm.
1: C: represents a counter
2: S: represents a state
3: procedure TRAFFIC REGULA-
TION (C,S,EN,ES,EE, EO, SN, SS, SE,SO)
4: if (EE + EO) 6= 0 then
5: rap ← (EN + ES)/(EE + EO)
6: else
7: rap ← 100
8: if (C < MAX and rap > 1.5) or C < MI N then
9: C ← C + 1
10: GreenLight(NSAll)
11: if S 6= NSAll then
12: C ← 0
13: S ← NSAll
14: else if (C < MAX and rap < 0.66) or C < MIN
then
15: C ← C + 1
16: GreenLight(EOAll)
17: if S 6= EOAll then
18: C ← 0
19: S ← EOAll
20: else if C = MAX or C = MIN then
21: C ← 0
22: if S = NSAll then
23: GreenLight(EOAll)
24: S ← EOAll
25: else if S = EOAll then
26: GreenLight(NSAll)
27: S ← NSAll
3.2 Hacking Scenario’s
Denial of Service (DoS) consists in making differ-
ent resources and services for users in the network
unavailable; it is usually caused by other attacks on
bandwidth or energy resources of other vehicles. It
is necessary to supervise and detect DoS in order to
avoid an incident. DoS attacks can lead to abnor-
mal conditions, preventing, intercepting or blocking
communication between vehicles in a VANET. In our
case, we consider the CCC as an important unit which
is able to attribute services to all vehicles for one in-
tersection.
Figure 2 depicts a service demand by a malicious
vehicle m. The m asks several services to the CCC
which makes the CCC busy. When a vehicle v asks
for one service, the CCC denies the requested service.
To resolve this problem, we delimited a number of
requested services. Each vehicle can ask one service
at a time.
CCC
m
v
Service demand
broadcast
All service
attribution
Service demand
Service not available
(DoS)
Figure 2: Denial of Service using our scenario.
Figure 3: Two double-handed lanes and an intersection.
4 EXPERIMENTAL EVALUATION
Map Definition
The scenario used in the experiments has two roads.
Both roads share a single intersection and have four
lanes. In each road, two lanes are from South/East to
North/West and two are on the opposite flow. This
ICINCO 2018 - 15th International Conference on Informatics in Control, Automation and Robotics
422

Figure 4: Vehicles average speed for each direction with and without optimal traffic lights control using Matlab based on
SUMO simulations.
scheme is shown in Figure 3. To implement these
roads, we used SUMO, TraCI4Matlab and Matlab.
Experiments
The system simulations were done in Matlab en-
vironment. TraCI4Matlab is built on top of the
TCP/IP stack, implemented on Traffic Control Inter-
face (TraCI) application level protocol. It connects
the client MATLAB together with the server SUMO
so that client can access and modify the simulation
environment. TraCI4Matlab allows MATLAB to take
control of SUMO objects such as vehicles, traffic
lights, etc., providing users a testbed to evaluate traf-
fic lights control protocol or any other related traffic
algorithm.
We compare, in Figure 4, the vehicles average
speed on their roads for each direction (EN, ES, EE,
EO, SN, SS, SE and SO). This comparison is related
to two cases of study: traditional and optimal traffic
lights control. The first case of study is represented
in the continuous line, the second one in discontin-
uous line. The latter uses the developed algorithm
given in section 3.1. We replace the two values MIN
and MAX from the algorithm by 15 and 50 u.t (unit
of time). The statistical obtained results are shown
in Figure 4 where the simulation duration is around
17 min (1000 s). We can notice that the average speed
is higher in the optimal case than in the traditional one
for most of the directions.
Figure 5: Vehicles average speed simulation for the whole
itinerary with and without optimal traffic lights control.
To confirm this result, we performed a simulation
for the whole vehicles trajectory. Figure 5 shows the
obtained result. As we can see, the average speed in
the traditional case is 7m/s while in the optimal case
is 11m/s. This result is improving the vehicle aver-
age speed by around 4m/s. In other words, the ve-
hicle’s speed is improved by 36% compared to the
traditional one. Therefore, the optimal traffic lights
control greatly improves the road traffic.
Preliminary Results for Secure Traffic Regulation
423

5 CONCLUSIONS
In this paper we presented an algorithm to regulate
the traffic for one intersection in ITS. This devel-
oped algorithm is at the same time centralized within
a CCC which is in charge of controlling the traffic
lights and also distributed among the vehicles that are
in the intersection. We also introduce one hacking
scenario and our proposed resolution. In order to test
the proposed algorithm, we implemented it on Mat-
lab and connected with SUMO. TraCI4Matlab is built
on top of the TCP/IP stack, implemented on Traf-
fic Control Interface (TraCI) application level proto-
col. It connects the client MATLAB together with the
server SUMO so that client can access and modify the
simulation environment. The obtained results shown
that our implemented algorithm improves the vehicles
speed and regulate the traffic better than the classical
regulation traffic algorithm. This result is encourag-
ing and is pushing us to implement the approach on
robot cars.
REFERENCES
Behrisch, M., Bieker, L., Erdmann, J., and Krajzewicz,
D. (2011). SUMO: simulation of urban mobility
(an overview). In Proceedings of the Third Interna-
tional Conference on Advances in System Simulation
(SIMUL 2011), pages 63–68.
Chembe, C., Noor, R. M., Ahmedy, I., Oche, M., Kunda, D.,
and Liu, C. H. (2017). Spectrum sensing in cognitive
vehicular network: State-of-art, challenges and open
issues. Computer Communications, 97:15 – 30.
Eze, S. Z. E. C. and Liu, E. (2014). Vehicular ad hoc net-
works (VANETs): Current state, challenges, poten-
tials and way forward. In Proceedings of the 20th
International Conference on Automation and Comput-
ing, pages 176–181.
Ghena, B., Beyer, W., Hillaker, A., Pevarnek, J., and Hal-
derman, J. A. (2014). Green lights forever: Analyzing
the security of traffic infrastructure. In 8th USENIX
Workshop on Offensive Technologies (WOOT 14), San
Diego, CA. USENIX Association.
Hasrouny, H., Samhat, A. E., Bassil, C., and Laouiti, A.
(2017). Vanet security challenges and solutions: A
survey. In Vehicular Communications, Volume 7,
pages 7–20.
Karagiannis, G., Altintas, O., Ekici, E., Heijenk, G., Jaru-
pan, B., Lin, K., and Weil, T. (2011). Vehicular net-
working: a survey and tutorial on requirements, archi-
tectures, challenges, standards and solutions. In IEEE
Communications Surveys & Tutorials, 13 (4), pages
584–616.
Leiding, B., Memarmoshrefi, P., and Hogrefe, D. (2016).
Self-managed and blockchain-based vehicular ad-hoc
networks. In Proceedings of the 2016 ACM Interna-
tional Joint Conference on Pervasive and Ubiquitous
Computing: Adjunct, UbiComp ’16, pages 137–140,
New York, NY, USA. ACM.
Lim, K. and Manivannan, D. (2016). An efficient proto-
col for authenticated and secure message delivery in
vehicular ad hoc networks. In Vehicular Communica-
tions, vol. 4, pages 30–37.
Lin, R., Kraus, S., and Shavitt, Y. (2007). On the ben-
efits of cheating by self-interested agents in vehicu-
lar networks. In Proceedings of the 6th International
Joint Conference on Autonomous Agents And Multi-
agent Systems (AAMAS 2007), New York, NY, USA,
pages 327–334.
Lin, X. and Li, S. (2013). Achieving efficient cooperative
message authentication in vehicular ad hoc networks.
In IEEE Transactions on Vehicular Technology, 62 (7),
pages 3339–3348.
Rawat, D. B., Bista, B. B., Yan, G., and Olariu, S. (2014).
Vehicle-to-vehicle connectivity and communication
framework for vehicular ad-hoc networks. In 2014
Eighth International Conference on Complex, Intel-
ligent and Software Intensive Systems (CISIS), pages
44–49.
Sharma, P. K., Moon, S. Y., and Park, J. H. (2017). Block-
vn: A distributed blockchain based vehicular network
architecture in smart city. In Journal of Information
Processing Systems, vol. 13, no. 1, pages 184–195.
ICINCO 2018 - 15th International Conference on Informatics in Control, Automation and Robotics
424