Multi-Component V2X Applications Placement in

Edge Computing Environment

Ibrahim Shaer* , Anwar Haque†, and Abdallah Shami*
*
Department of Computer and Electrical Engineering,†Department of Computer Science
Western University, London ON, Canada
{ishaer, ahaque32, abdallah.shami}@uwo.ca

Abstract—Vehicle-to-everything (V2X) services are attracting [1] . Additionally, the centralized servers are usually located
arXiv:2005.06956v1 [cs.NI] 22 Apr 2020

a lot of attention in the research and industry communities far from vehicles generating data; thus, incurring a huge E2E
due to their applicability in the landscape of connected and delay. Furthermore, this architecture exposes a single point of
autonomous vehicles. Such applications have stringent perfor-
mance requirements in terms of complex data processing and failure, which is huge risk to take for time and mission-critical
low latency communications which are utilized to ensure road V2X applications. Given these circumstances, distributing the
safety and improve road conditions. To address these challenges, cloud computing technology in proximity to users is proposed
the placement of V2X applications through leveraging of edge as a viable solution to deal with the shortcomings of the
computing paradigm, that distributes the computing capabilities centralized paradigm [2]. This computing architecture is re-
to access points in proximity to the vehicles, presents itself
as a viable solution. However, the realistic implementation of ferred to as Edge Computing. Edge Computing can support
the edge enabled V2X applications is hindered by the limited the latency requirements of the V2X applications which are
computational power provided at the edge and the nature of critical for their performance [3]. In addition, the edge servers
V2X applications that are composed of multiple independent V2X collect data from the close local nodes which allows for
basic services. To address these challenges, this work targets the a more individualized experience for the V2X application
efficient placement of V2X basic services in a highway scenario
subject to the delay constraints of V2X applications using them users. While Edge Computing paradigm can ensure some
and the limited computational resources at the edge. To that end, V2X system-level performances, this comes at the expense
this work formulates a binary integer linear programming model of limited computational power at the edges which hinders
that minimizes the delay of V2X applications while satisfying the the processing of large amount of data. Microservices archi-
resource requirements of V2X basic services. To demonstrate tecture, that decomposes a single application into decoupled
the soundness of the approach, simulations with varying vehicle
densities were conducted, and the results reported show that it modules, combined with virtualization techniques, that fully
can satisfy the delay requirements of V2X applications. utilizes resource at the edges, can be used to address this
Index Terms—V2X applications, edge computing, placement, issue. Hawilo et. al [4] investigated the applicability of this
multi-component paradigm for Virtual Network Functions which display similar
characteristics to V2X applications making it a viable option
I. I NTRODUCTION for their placement. In the domain of V2X applications, 3rd
Intelligent Transportation Systems (ITSs) are envisioned to Generation Partnership Project (3GPP) [5] envisions complex
ameliorate traffic congestion and improve road safety and V2X applications that combine vehicle status analysis, immi-
traffic experience. ITSs have drawn the attention of a large nent traffic events generation, and raw sensor data exchange
number of stakeholders due to their direct effect on the man- that define the function of autonomous and connected vehicles.
ufacturing of sensor and wireless-equipped vehicles known as Each of these applications rely on the data processing and
connected and autonomous cars. In this regard, Vehicle-to- analysis of miniscule V2X basic services.
everything (V2X) applications are considered a key enabler for Mobile edge clouds (MEC), edge clouds and roadside cloud
the shift to Intelligent Transportation Systems (ITSs) in terms have been proposed in several previous works in the context
of traffic management. These applications allow the vehicles to of vehicular applications. In [6], Emara et. al employ an
communicate and exchange information with their surrounding MEC-assisted architecture to evaluate end-to-end latency for
environment that includes other vehicles, pedestrians and vehicles to detect the vulnerable road units. Moubayed et.
supporting road side units (RSUs). To ensure road safety, al [7] formulated an integer linear programming problem
these applications operate with stringent end-to-end (E2E) for efficient placement of V2X basic service taking into
latency/delay. There are different paradigms that can determine consideration V2X basic services’ delay and computational
the placement of V2X applications to address the E2E latency requirements in a hybrid environment that includes edge and
requirements. The placement of these services is disruptive core nodes. Supporting V2X applications while considering
to the customary cloud-based infrastructure. The projected the vehicle’s mobility aspects has been extensively addressed
increase in the number of connected and autonomous vehicles in literature. To support V2X applications, [8]–[10] consider
will result in data explosion. The data will be routed to a single migrating the services according to vehicle’s mobility. In [8],
centralized server creating severe network traffic congestion the authors customize a three-layered architecture that consists
of a vehicular cloud, a roadside cloud and a central cloud Decentralized Notification Messages (DENM) whenever a
to support vehicular applications. Their approach focuses on road hazard or abnormal traffic condition takes place, and
the dynamic allocation of resources, driven by the vehicle’s Media Downloading [14] service is requested on demand
mobility, in vehicular and roadside clouds. In [9], Yu et. al by the passengers of the vehicle. Additionally, ETSI defines
consider the migration of V2X applications placed on edge Local Dynamic Maps (LDMs) [15] that are responsible for
servers according to predictive vehicle’s mobility combined storing the sent CAMs and DENMs. Because LDMs store
with setting a priority schema for V2X applications. The spatial relevant information, an LDM is deployed on each
approach considers the latency and resource requirements of edge server. LDMs are queried by V2X basic services in order
each of the applications. In the same context, Yao et. al [10] to retrieve information. Finally, in addition to basic vehicular
investigate Virtual Machine (VM) placement and migration in services that are related to road safety, there is a variety of
roadside cloud that is part of the vehicular cloud computing innovative applications that are referred to as value-added
architecture. The approach targets minimizing the overall services that are of lower priority [16]. These services include
network cost given the available resources at the edge. Each of augmented reality, parking location and others that are part of
these previous works has its own shortcoming. One common the infotainment services provided by vehicular applications.
aspect is considering either latency or resource limitations for Compared to road safety applications, these services display
the placement of the services at the edge. Another shortcoming high levels of diversity and individuation. Therefore, they need
is the disregard of the nature of V2X applications that may to be migrated when the vehicle moves from one edge server
be composed of a single or many modules. Finally, different coverage zone to another. For this purpose, each edge server
traffic conditions were not considered to model any solution reserves part of its resources to accommodate these migrating
for vehicular application placement. To address these short- services. In this section, the system design and the optimal
comings, this work focuses on V2X application placement that optimization technique for V2X basic service placement are
minimizes the end-to-end delay which takes into consideration presented.
the computational requirements of V2X services forming it.
This work’s main contributions are as follows: A. System Design
• Decompose V2X applications into multi-V2X basic ser- In the reference model used for the placement of V2X
vices. basic services, HWY 416 IC-721A that passes through the
• Formulate the optimal V2X application placement by city of Ottawa is considered. The edge computing servers are
considering their delay requirements and the resource deployed uniformly along the highway as the deployment of
requirements of their constituent components. RSU is out of the scope of this paper. No communication
• Evaluate the performance of the optimal placement in interference zone exists between any two successive RSUs to
terms of average delay and density distribution for each avoid the possibility of encountering ping-pong handover cases
V2X application under different traffic conditions. which will be difficult to handle in an optimization model.
The remainder of this paper is organized as follows: Section Additionally, the vehicles are assumed to be always connected
II describes the system model and presents the problem to RSUs throughout their journey. The end-to-end latency
formulation, Section III provides the simulation procedure and of a service is the sum of the communication, processing,
discusses the results and Section IV concludes the paper and transmission and propagation delay. The propagation delay is
suggests future work. dependent on the medium of communication which is out of
the scope of this paper, and therefore considered negligible. In
II. S YSTEM M ODEL this model, DSRC, the communication technology between the
In the reference model, a highway scenario is considered. moving vehicles and RSUs, affects the communication delay.
Each of the vehicles moving on the highway is running a set of In the proposed model, the processing and transmission delay
V2X applications that are collecting data from nearby roadside between the communicated edge servers is considered. Each
units (RSUs) to function autonomously. RSUs and the vehicles edge computing server has the same computing and processing
are communicating directly using Dedicated Short-range Com- power that are expressed by the number of cores and RAM
munication [11], and no communication takes place between available. Finally, the vehicle density is considered to model
any vehicles. Each RSU is equipped with a server which real case scenarios.
are both considered as an edge computing node. Vehicles are
receiving data from V2X basic services placed on each RSU. B. Optimization Problem
European Telecommunication Standardization Institute (ETSI) In the optimization function, a set of edge servers and
defines three V2X basic services that are the foundation of any V2X services are considered. Let N denote the set of edge
envisioned V2X applications. The V2X basic services are as servers where n ∈ N. Let U denote the set of unique
follows: Cooperative Awareness Basic Service (CA) [12] is V2X basic services where u ∈ U. The availability of the
responsible for creating, analyzing and sending Cooperative computational resources on the edge is denoted by matrix
Awareness Basic Messages (CAMs) which include informa- Cap where Capkn denotes the kth computational resources
tion about the vehicle’s status and attributes, Decentralized available on edge server n. Matrix R represents the resources
Environmental Notification Service [13] (DEN) broadcasts required by the V2X basic services where Rku represents the
kth computational resources required by V2X basic service u. Fig. 1. System Model
A binary row vector − →
q denotes the edge servers a vehicle can
communicate with. Let C be the matrix that represents the
processing and the transmission latency between edge servers
where Cij represents the latency between edge server i and
edge server j. Matrix M represents the V2X services needed
by V2X applications where Mau = 1 denotes that application
a needs V2X basic service u. Let X denote the placement
matrix where Xun = 1 means that V2X basic service u
is placed on edge server n. The column Xu denotes the
placement of the V2X services on edge server n. Dav and
Dath denote respectively the delay experienced by a moving
vehicle v served by application a and the maximum tolerable denoted by VM 3. The logic governing the realization of the
threshold of this delay. To represent the vehicles’ density, γ is application is as follows:
used. dvcom and dvDL denote respectively the communication (1) The vehicle requests the services of an application. This
and download latency between a vehicle v and a serving edge step incurs communication delay that is denoted by dvcom .
server. The optimization function used to minimize the delay (2) The CA service found on Edge Server 1 requests the
of V2X application is as follows: necessary information from the LDM. The processing of the
X request on the LDM is denoted by C11 .
min Dav (1) (3) Edge server 1 communicates with the closest server that
a∈A include DEN basic service. No delay is considered in this
where: phase.
(4) DEN queries and receives information from the LDM
that is closest to the requesting vehicle. The LDM on edge
Dav = dvcom + max(Ma ⊙ min(X ⊙ (γC × −
→
q )) + dvDL (2) server 1 has accurate information about the requesting ve-
subject to: hicle’s surrounding environment. This delay is the sum of
the processing delay of LDM on edge server 1 and the
Dav ≤ Dath , ∀a ∈ A (3) transmission delay between edge server 1 and 2. This is
denoted by C12 .
(5, 6) These steps represent the CA and DEN response to
RX ≤ Cap (4) the requesting vehicle. This delay is denoted by dvDL .
For basic service CA, the delay is as follows:
X
Xn = 1, ∀n ∈ N (5)
dvCA = C11 + dvDL
In what follows, the explanation of the equations (1)-
Similarly, the delay for DEN is:
(5). Equation (1) describes the overall objective which is
minimizing the summation of the delay of all V2X application
dvDEN = C12 + dvDL
experienced by a vehicle requesting their services. Equation
(2) presents the components contributing to the delay of V2X Given that the requests for each basic service are executed
application. The delay of a V2X application is the delay of in parallel and that these services are independent in their
the V2X services it relies on depending on the edge servers execution, the delay experienced by a vehicle v requesting the
a vehicle can communicate with. Because the functioning services of application a is:
of a V2X basic service is independent of other V2X basic
services, the delay of a V2X application is defined as the Dav = dvcom + max{dvCA , dvDEN }
maximum of the delay of its constituent V2X basic services.
III. E XPERIMENTAL S ETUP AND R ESULTS
This value is added to the communication and download link
delay. Next, the equations (3) – (5) describe the constraints. A. Simulation Setup
Equation (3) defines that the delay of an application should In order to evaluate the placement of V2X basic services,
not exceed its maximum defined tolerable delay. Equation a realistic simulation environment must be created. To this
(4) ensures that the resources allocated to a V2X service do end, Simulation of Urban Mobility (SUMO) [17] was used
not exceed the available resource on the hosting edge server. to extract the movement of vehicles along a highway. A 4
Equation (5) limits placing only one V2X service on each km highway that resembles HWY 416 IC-712A was con-
edge server. The following diagram illustrates an example of sidered as a reference highway. Ontario traffic volume for
the communication and processing that takes place for a V2X provincial highways [18] provided the average daily traffic
application that requires CA and DEN services, given that and the accident rates during summer, winter, weekdays and
each server has resources reserved for migrating applications weekends. In the simulation setup, the statistics offered by
 TABLE I TABLE II
SUMO VEHICLE MOVEMENT PARAMETERS B REAKDOWN OF V2X APPLICATIONS ’ V2X BASIC SERVICE AND
PERFORMANCE METRICS
Parameter Value
Maximum Speed 27.7m/s Application Service(s) Latency(ms) Reliability(%)
Maximum Acceleration 2.6m/s2 PL CA 50 90
Maximum Deceleration 4.5m/s2 SSM CA, DEN, Media 20 90
ES DEN 10 95
PSW CA, DEN 20 95
FCW CA, DEN 10 95
this report were used to emulate moderate and heavy traffic
experienced on HWY 416 IC-712A highway that is expressed TABLE III
through the vehicles per hour parameter in SUMO. Regarding C OMPUTATIONAL R EQUIREMENTS
the movement of the vehicles, Table I summarizes the key
Entity Number of Cores RAM
parameters that are used in the simulation.
Edge Server 8 8
The V2X applications considered are Platooning (PL), CA 2 2
Sensor and Sensor State Mapping (SSM), Emergency Stop DEN 2 4
(ES), Pre-crash Sensing Warning (PSW) and Forward Colli- Media Service 4 6
LDM 4 2
sion Warning (FCW). Their corresponding performance re-
quirements and service components are presented in Table
II [19], [20]. Choosing these V2X applications stems from CA, DEN and Media services. In the experimental procedure,
their importance and stringent performance requirements in the placement of the V2X basic services is carried out using
the realm of the autonomous cars. In addition, in the context the defined optimization function. Next, traffic simulation is
of the defined problem, each of the chosen V2X application executed for defined densities that reflect moderate and heavy
offers a unique combination of V2X services. In the simulation traffic. The traffic traces were generated for 1500 seconds.
procedure, the communication delay between a vehicle and an Every 10 seconds, a snapshot of the road condition is taken and
RSU is 1 ms [21]. In this model, the processing delay is the delays for each V2X application for each vehicle is calculated.
amount of time required by a Local Dynamic Map to process Finally, at the end of the simulation, the average delay for each
the data requested by other V2X services either placed on the V2X application is obtained.
same or different edge server. In [22], the authors devise an
LDM according to the specifications defined by ETSI. The B. Implementation
application defines two Application Programming Interfaces The optimization function was solved using IBM ILOG
(APIs) that retrieve information of the IDs of the vehicles CPLEX 12.9.0 through its Python API. The solution is pro-
driving on the same road and the vehicle driving immediately vided instantly for all simulation scenarios with different
ahead of the requesting vehicle. For different number of vehicle densities on a laptop with an Intel Core i7-8750 CPU,
queried vehicles ranging from 5 to 20 vehicles, the response 2.21 GHz clock frequency and 16 GB of RAM. The final
time was between 3 and 5 ms with no clear correlation between solution includes the V2X services placed on each edge server.
the size of the data and the response time. Consequently, in the
simulation setup, the processing delay is generated uniformly C. Results and Discussion
between 3 and 5 ms. In the same context, the authors in [23], To evaluate the efficacy of the optimization function, the
assumed the transmission latency between two edge servers simulation procedure was carried out using two different traffic
to be between 1 and 5 ms. Because the simulation procedure scenarios each representing moderate (Scenario 1) and heavy
takes place under several vehicle densities, the increase of the (Scenario 2) traffic models. The results are obtained as an
data processing and transmission overhead with the increase of average for five independent runs. To assess the placement
number of vehicles is inevitable. In this regard, the execution function, the average delay of each V2X application under
cost increases with the number of vehicles in proximity to study is obtained and compared it to the maximum toler-
the vehicle requesting the V2X application services. As the able delay. Additionally, the model is evaluated using the
implementation of LDM did not consider cases beyond 20 probability density function of each of V2X application. The
vehicles, the added delay for these cases will be in the form density function provides a more thorough overview about
of log(N C/20) where N C represents the number of cars and the distribution of the delays in terms of detecting extreme
the expression is derived from the increase of processing delay values that are overshadowed by the common values trend.
upon the increase in size of the queried data in SQL [24]. In Furthermore, the density functions reveal the shortcomings of
terms of edge servers, 10 edge servers are deployed every 400 the approaches that are concealed by the calculation of the
m alongside the highway. Each of the RSUs hosts an LDM, mean. The suggested optimization function failed to converge,
V2X service and an optional migrating service. The computa- so a new heuristic algorithm that relaxes the delay threshold
tional requirements of CA, DEN and Media services are those for each application by magnitudes of the reliability metrics is
of a small, medium and large VMs. Table III summarizes the considered and executed. This heuristic algorithm is referred
edge server capabilities and the computational requirements of to as: Resource and Delay-aware V2X basic service Placement
Fig. 2. Average Delay of V2X Applications for different Vehicle Densities Fig. 4. Probability Density Function for 1800 vehicles/hour
The delay experienced by V2X applications for different vehicle densities Probability Density Function of PL Probability Density Function of SSM
0.25 0.3
12 Scenario 1
Scenario 2
0.25
0.2

10 0.2
0.15
0.15

8 0.1
0.1
Delay (ms)

0.05
0.05
6
0.0 0.0
4.0 5.0 6.0 7.0 8.0 9.0 10.0 8.0 10.0 12.0 14.0 16.0 18.0
Delay (ms) Delay (ms)
4 Probability Density Function of ES Probability Density Function of PSW
0.25
0.4
2 0.2

0.3
0.15
0
PL

SSM

ES

PSW

FCW
0.2
0.1

0.1
0.05

0.0 0.0
Fig. 3. Probability Density Function for 1500 vehicles/hour 4.5 5.0 5.5 6.0
Delay (ms)
6.5 7.0 7.5 8.0 8.5 5.0 6.0 7.0
Delay (ms)
8.0 9.0 10.0

Probability Density Function of PL Probability Density Function of SSM Probability Density Function of FCW
0.4 0.3
0.25
0.35
0.25
0.3 0.2
0.2
0.25
0.15
0.2 0.15

0.15 0.1
0.1
0.1
0.05 0.05
0.05

0.0 0.0 0.0

4.0 5.0 6.0 7.0 8.0 9.0 10.0 8.0 10.0 12.0 14.0 16.0 5.0 6.0 7.0 8.0 9.0 10.0
Delay (ms) Delay (ms) Delay (ms)
Probability Density Function of ES Probability Density Function of PSW
0.4

0.4 0.35

0.3
0.3
0.25 for FCW application which shows that for each scenario,
0.2
0.2
0.15
20% and 25% of the experienced delay exceeds the tolerable
0.1
0.1 threshold which is beyond the 5% permitted shown in Table
0.05

0.0
4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5
0.0
5.0 6.0 7.0 8.0 9.0 10.0
II. In terms of traffic effect, it is observed that there is a
Delay (ms) Delay (ms)
Probability Density Function of FCW slight right shift of the probability distribution in scenario 2.
0.4

0.35
Additionally, it is observed that some applications have similar
0.3
probability distributions. This is attributed to the fact that these
0.25

0.2
applications need the same V2X services, and as it shows,
0.15
these services incur the most delay out of the other services
0.1

0.05
that they rely on. The dispersion of some of the probability
0.0
5.0 6.0 7.0
Delay (ms)
8.0 9.0 10.0 density function is due to the limited number of edge nodes
hosting V2X services. The limited number of edge servers
means that vehicles at the start and the end of the route will
(RDP). The results of the simulation process in terms of suffer from prolonged delay due to the distance separating
the average delay and the probability densities of each V2X the vehicles and the closest V2X basic services. In the cases
application are presented in Figures 2-5. of continued route, the suggested approach can be replicated
Figure 2 shows the mean delay for each of the V2X along the highway to ensure that V2X services are delivered
applications. The results clearly show that the average delay as expected. For comparison purposes and to further cement
experienced by each V2X application is within the tolerable this paper’s approach, a baseline approach that maximizes
threshold. These results show that the heuristic algorithm met the resource utilization at each node server is compared to
the stringent V2X application delay requirements. In terms RDP. The baseline approach formulates a placement algorithm
of the traffic effect, the mean of delay of each application that takes into consideration only the available resources at
has slightly increased but still fulfills the overall objective of each node. This baseline approach is reffered to as Resource-
the placement function. The vehicles’ density has contributed Aware Algorithm (RAA). The two approaches were evaluated
to an increase in the average delay of the V2X applications according to the probability density functions of the delays
in the range of 1.3% to 4.8% whereby the SSM application of ES and FCW applications. The probability densities are
has experienced the greatest variation. This fact shows that depicted in Figure 5.
the placement of Media Service, that SSM relies on, is the The baseline line approach’s density function shows promis-
most sensitive to traffic variation. On the other hand, the ing results regarding the ES application as the full delay
probability density functions tell a different story. Figures 3 distribution is below the tolerable threshold. More values are
and 4 depicting the delay distribution for both cases show concentrated on the extremes which makes it harder to gauge
that the delay is highly skewed to the left which supports its value whenever the application is requested. However, for
the viability of the approach. However, this is not the case the case of FCW, this approach fails to be within the tolerable
 Fig. 5. RDP vs RAA [7] A. Moubayed, A. Shami, P. Heidari, A. Larabi, and R. Brunner, “Edge-
The delay for ES Application The delay for FCW Application enabled V2X Service Placement for Intelligent Transportation Systems.”
RDP 0.2 RDP
0.4 RAA RAA in IEEE Transactions on Mobile Computing.
0.18
0.35 [8] R. Yu, Y. Zhang, S. Gjessing, W. Xia, and K. Yang, “Toward cloudbased
0.15
0.3 vehicular networks with efficient resource management,” IEEE Netw., vol.
0.25 0.12
27, no. 5, pp. 48–55, Sep./Oct. 2013.
0.2 0.1
[9] X. Yu, M. Guan, M. Liao, and X. Fan, “Pre-migration of vehicle to
0.15 0.08

0.05
network services based on priority in mobile edge computing,” IEEE
0.1

0.05 0.02
Access, vol. 7, pp. 3722–3730, 2019.
0.0 0.0
[10] H. Yao, C. Bai, D. Zeng, Q. Liang, and Y. Fan, “Migrate or not? explor-
5.0 6.0 7.0 8.0 9.0 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5
Delay (ms) Delay (ms) ing virtual machine migration in roadside cloudlet-based vehicular cloud,”
Concurrency and Computation: Practice and Experience, 27(18):5780–
5792, 2015.
[11] Z. Xu, X. Li, X. Zhao, M. H. Zhang, and Z. Wang, “DSRC versus
threshold rendering this approach ineffective for mission- 4G-LTE for Connected Vehicle Applications: A Study on Field Experi-
critical applications. This is to be expected given that FCW ments of Vehicular Communication Performance,” Journal of Advanced
application requires CA and DEN basic services. Due to the Transportation, vol. 2017, pp. 1–10, 2017.
[12] ETSI EN 302 637-2 (V1.3.2), “Intelligent Transport Systems (ITS) Ve-
nature of RAA that maximizes the overall resource utilization, hicular Communications: Basic Set of Applications - Part 2: Specification
deploying more of CA services results in decreasing the of Cooperative Awareness Basic Service,” 650 Route des Lucioles F-
utilization which incurs extra delay for FCW application when 06921 Sophia Antipolis Cedex - FRANCE, Technical Report., 2014.
[13] ETSI EN 302 637-3 (V1.2.1), “Intelligent Transport Systems (ITS)
requesting the services of CA. Vehicular Communications: Basic Set of Applications - Part 3: Spec-
ifications of Decentralized Environmental Notification Basic Service,”
IV. C ONCLUSION 650 Route des Lucioles F- 06921 Sophia Antipolis Cedex - FRANCE,
This paper addressed the efficient placement of V2X basic Technical Report, 2014.
[14] ETSI TR 102 638 (V1.1.1), “Intelligent Transport Systems (ITS) Ve-
service comprising different V2X applicatins in an edge com- hicular Communications: Basic Set of Applications - Definitions,” 650
puting environment. To this end, an optimization function that Route des Lucioles F-06921 Sophia Antipolis Cedex - FRANCE, Tech.
minimizes the delay for multi-component V2X applications Rep., 2009.
[15] ETSI TR 102 863 (V1.1.1), “Intelligent Transport Systems (ITS);
consisting of V2X services while considering the resource Vehicular Communications; Basic Set of Applications; Local Dynamic
requirements of these services under different traffic conditions Map (LDM); Rationale for and Guidance on Standardization”, 650 Route
is formulated. The approach was evaluated under realistic des Lucioles F- 06921 Sophia Antipolis Cedex - FRANCE, Technical
report, 2011.
scenarios where homogeneous edge servers with limited com- [16] L. Li, Y. Li, and R. Hou, “A Novel Mobile Edge Computing-Based
putational power and variable traffic conditions were consid- Architecture for Future Cellular Vehicular Networks,” 2017 IEEE Wireless
ered. Furthermore, the approach was compared to a base- Communications and Networking Conference (WCNC), 2017.
[17] D. Krajzewicz, J. Erdmann, M. Behrisch, and L. Bieker, “Recent
line approach that maximizes the overall resource utilization development and applications of SUMO - Simulation of Urban MObility,”
of edge servers. The results have shown that the approach International Journal On Advances in Systems and Measurements, vol.
guarantees an acceptable quality of service, and outperforms 5, no. 3&4, pp. 128–138, December 2012.
[18] Ontario Ministry of Transportation, “Provincial Highways Traffic
other approaches while emulating realistic conditions. While Volumes 1988-2016,” December 31, 2016. [Online]. Available:
the current work considers that each V2X application has a http://www.raqsa.mto.gov.on.ca/techpubs/TrafficVolumes.nsf/fa02780864
constant request rate, the plan is to extend the work to consider 7879788525708a004b5df8/88c66a2279555c798525788d0048cca4/$FILE
/Provincial%20Highways %20traffic%20Volumes%201988-2016.pdf.
different request distributions to mimic a real-world scenario. [19] 3GPP, “Study Enhancement 3GPP Support for 5G V2X Services (Re-
In the same context, the deployment of V2X applications in a lease 15),” TR 22.886 V15.1.0, March 2017.
dynamic service availability environment is also a subject to [20] 3GPP, “Study Enhancement 3GPP Support for 5G V2X Services (Re-
lease 16),” TR 22.886 V15.1.0, December 2018.
our future work. [21] D. Martin-Sacristan, S. Roger, D. Garcia-Roger, J. F. Monserrat, A.
Kousaridas, P. Spapis, S. Ayaz, and C. Zhou, “Evaluation of lte- advanced
R EFERENCES connectivity options for the provisioning of v2x services,” in 2018 IEEE
[1] Barrachina, J., Garrido, P., Fogue, M., Martinez, F., Cano, J., Calafate, Wireless Communications and Networking Conference (WCNC), Apr.
C., and Manzoni, P, “Road Side Unit Deployment: A Density-Based 2018, pp. 1–6.
Approach”. IEEE Intell. Transp. Syst. Mag. 2013, 5, 30–39. [22] Shimada H. , Yamaguchi A., Takada H. , and Sato K., “Implementation
[2] Faticanti F., De Pellegrini F., Siracusa D., Santoro D. , and Cretti S., “Cut- and Evaluation of Local Dynamic Map in Safety Driving Systems,”
ting throughput on the edge: App-aware placement in fog computing”. Journal of Transportation Technologies, vol. 05, no. 02, pp. 102–112,
CoRR, abs/1810.04442, 2018. 2015.
[3] Sabella D., Moustafa H., Kuure P., Zhou Z., Li A. et al. “Toward [23] Maheshwari, S.; Raychaudhuri, D.; Seskar, I.; Bronzino, F., “Scalability
fully connected vehicles: Edge computing for advanced automotive and Performance Evaluation of Edge Cloud Systems for Latency Con-
communications ,” 5GAA, December 2017. [Online]. Available: strained Applications,” In Proceedings of the 2018 IEEE/ACM Symposium
https://5gaa.org/wp-content/uploads/2017/12/5GAA T-170219- on Edge Computing (SEC), Seattle, WA, USA, 25–27 October 2018; pp.
whitepaper-EdgeComputing 5GAA.pdf. 286–299
[4] Hassan Hawilo, M. Jammal and A. Shami, "Exploring Microservices [24] “MySQL 8.0 Reference Manual::13.8.2 EXPLAIN Statement,”
as the Architecture of Choice for Network Function Virtualization Plat- MySQL. Accessed on: Oct. 19, 2019. [Online]. Available:
forms," in IEEE Network, vol. 33, no. 2, pp. 202-210, March/April 2019. https://dev.mysql.com/doc/refman/8.0/en/explain.html.
[5] Study LTE Support for Vehicle to Everything (V2X) Services (Release 14),
document 3GPP TR 22.885 V14.0.0, Dec. 2015.
[6] M. Emara, M. C. Filippou, and D. Sabella, “MEC-Assisted End-to-End
Latency Evaluations for C-V2X Communications,” in 2018 European
Conference on Networks and Communications (EuCNC), Jun. 2018, pp.
1–9.