Proposal of an Automotive Collision Avoidance System based on

Edge Computing

Nicolas Nevigato, Mauro Tropea and Floriano De Rango

Dimes Department, University of Calabria, Via P. Bucci 39/c, 87036 Rende (CS), Italy

Keywords:

Automotive, Edge Computing, Multi-Access Edge Computing (MEC), Collision Avoidance.

Abstract:

Multi-Access Edge Computing (MEC) is a network architecture that enables cloud computing capabilities and

an IT service environment at the edge of the cellular network and, more in general at the edge of any network.

It is often used in safety-critical and mission-critical communications as it is able to guarantee low latency

and high bandwidth. It takes advantage of container-based virtualization and in collaboration with 5G cellular

networks, it is a key enabler for the deployment of vehicular use cases, such as the improvement of road

safety. This paper presents the implementation of a collision avoidance system based on MEC, with the ability

to switch communication to a cloud server when possible, trying to guarantee the required constraints and

balancing the communication among the servers avoiding of overloading edge layer. The simulation results

have proved how in some cases the MEC-5G combination is the best one for the system in order to avoid

collisions on the road.

1 INTRODUCTION

The automotive industry is constantly evolving,

thanks to the numerous and important technologi-

cal transformations it is going through (Giust et al.,

2018). Internet of Vehicles (IoVs) (Yang et al.,

2014) is a network of cars communicating with each

other and with pedestrians handheld devices in a dis-

tributed manner supporting the use of data created

by connected cars and Vehicular Ad-Hoc Networks

(VANETs) (Sharma et al., 2019). Vehicles are be-

coming smarter and more aware of their surroundings

through the integration of new sensors and new com-

munication systems. They are able to exchange in-

formation with other vehicles (V2V communication),

with road infrastructure (V2I communication), with

pedestrians (V2P communication) and with commu-

nication networks (V2N communication). All these

types of communication are incorporated in the so-

called Vehicle-to-Everything communication (V2X

communication), that is, communication from vehicle

to everything.

The vehicular network aims to improve road

safety, increase trafﬁc efﬁciency and save energy giv-

ing also attention to the emission issues (Santamaria

et al., 2019), (Fazio et al., 2017b). Very studied as-

pects for the ad-hoc network are routing issues as it

is possible to view in (Zhou et al., 2006), (Fotino

Figure 1: Communications in vehicular networks.

et al., 2007), (Socievole et al., 2011), and also for

VANET is an important topic as it is possible to view

in (De Rango et al., 2009), (Fazio et al., 2013). The

cell permanence time and the mobile analysis for pre-

dictive services are important aspects in wireless net-

works and they are object of study by research com-

munities, such as it possible to view in (De Rango

et al., 2008), (Fazio et al., 2017a), (Fazio et al., 2016),

(Frnda et al., 2013). A fundamental aspect to con-

stantly consider in the automotive context is the pos-

sibility of guaranteeing an effective service in order

to obtain low-latency communications. The 5G net-

work can help in this direction, as it guarantees ultra-

low latency and ultra-high reliability in highly mo-

bile and densely connected scenarios. In addition,

the Multi-access Edge Computing, also called Mo-

128

Nevigato, N., Tropea, M. and De Rango, F.

Proposal of an Automotive Collision Avoidance System based on Edge Computing.

DOI: 10.5220/0009971101280135

In Proceedings of the 10th International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH 2020), pages 128-135

ISBN: 978-989-758-444-2

bile Edge Computing, (MEC) network architecture

(Porambage et al., 2018), (Pham et al., 2019), which

brings the processing, storage and network capabil-

ities to the edge of the network, can be considered

a key factor in these vehicular networks. In fact, it

meets the requirements of low latency and high band-

width and can be implemented on the roadside and

in the infrastructures of mobile network operators to

support vehicle-to-infrastructure communication.

Several applications represent a concrete use case

of MEC technology in collaboration with the latest

generation networks. A possible use case in the au-

tomotive sector is to make a driver’s driving safer,

through assisted driving systems to improve the re-

liability, efﬁciency and quality of road transport. For

example, it is possible to think of a system capable

of predicting potential collisions by anticipating de-

tection of a road hazard. In aid of this, it is possible

to use virtualization (Jain and Choudhary, 2016) with

Docker containers (Bernstein, 2014) and the container

orchestrator Kubernetes (Shah and Dubaria, 2019).

The rest of this paper is organized as follows: Sec-

tion 2 presents a brief overview on the vehicular en-

vironment; Section 3 presents the proposal of this

work on the considered research topic. In Section 4,

a description of the developed system is given with a

description of the simulation environment created in

java and the simulator implementation details. The

numerical results are presented in Section 5. Finally,

Section 6 concludes the paper.

2 RELATED WORK

A lot of work exist in literature on topics that regard

the automative envirnoment and Vehicular Ad-Hoc

Networks (VANETs). In general, they face with dif-

ferent issues such as communication protocols, new

technologies used in particular scenarios, virtualiza-

tion capabilities and so on. Many works have been

written on the new paradigm called Internet of Vehi-

cles (IoVs).

In (Song and Jiang, 2015) the authors aim at dis-

covering the related characteristics between the ac-

tive safety strategy and velocity through the visual

analysis of Internet of Vehicles (IoV) warning data.

They propose a warning velocity association model

(WVAM) able to visually validate the relation be-

tween warnings and velocity. Experiments using

Gephi show meaningful results.

In (Xiong et al., 2019) the authors propose an

elastic wave equation model to reveal the relation be-

tween safety distance and road congestion. It can be

found that the propagation speed of road congestion is

largely affected by the safety distance. To recoup the

efﬁcient road safety and alleviate road congestion, an

optimization problem is formulated with cooperative

communication and computing via platoons that aims

to minimize the total safety distance. Simulation ex-

periments show that the proposed algorithm leads to

near-optimal results with low complexity but no over-

head of vehicular information exchange.

In (Khelil and Soldani, 2014) the authors explore

recent Device-to-Device (D2D) research efforts and

review their suitability to safety-critical Internet of

Vehicles (IoV) applications, such as cooperative or

autonomous driving. Their review work shows that

current approaches ignore high relative node mobility

and accurate proximity measurements, and the crucial

research challenges mainly are those related to main-

taining the required QoS level in highly ﬂuctuating

D2D communications. In this work the authors want

to provide a roadmap towards adopting the emerging

D2D technique for critical IoV communications.

In (Santamaria et al., 2018) the attention is fo-

cused on the design of a new approach for vehicular

environments able to gather information during mo-

bile node trips, for advising dangerous or emergency

situations by exploiting on-board sensors. It is as-

sumed that each vehicle has an integrated on-board

unit composed of several sensors GPS device, able to

spread alerting messages around the network, regard-

ing warning and dangerous situations/conditions and

it is assumed that each vehicle can communicate with

on roadside devices called RSU able to spread warn-

ing messages. In this way, if an accident occurs, the

arriving cars will, probably, avoid delay and danger

situations.

3 PROPOSED SOLUTION

The system implemented for the use case shown

above has the following operating principle: the vehi-

cles periodically, by a device installed on board, send

messages that contain information about their status.

In particular, the message contains the ID, the posi-

tion and speed of the vehicle in question. These mes-

sages are received by a MEC server connected to a

roadside base radio station, running on a Docker (doc,

2019) container managed by Kubernetes (kub, 2019).

The information is saved and processed by a software

that can understand when an emergency braking is

taking place and therefore anticipate any collisions.

In this case the MEC server generates an alert mes-

sage, and immediately sends it to the following vehi-

cles within a certain range. When the alert message

reaches the vehicles, automatically, by means of suit-

Proposal of an Automotive Collision Avoidance System based on Edge Computing

129

able devices, the braking system is activated to stop

the vehicle safely and avoid a collision.

A dynamic switching algorithm has also been im-

plemented that allows the vehicle to instantly decide

whether to send the message to the MEC server or to

a cloud located in Europe, so as not to overload the

device on the roadside when it is not necessary.

3.1 Docker and Kubernetes

A container is an isolated environment that shares the

same kernel of the operating system. It contains the

code and all the dependencies to be able to run an

application quickly and easily.

Docker is an open source project that provides a

systematic way to automate the faster deployment of

Linux applications within portable containers (Bern-

stein, 2014; Ahmed and Pierre, 2018). It uses a client-

server architecture model and is based on the concept

of image. A Docker image can include only the fun-

damentals of the operating system, or it can consist of

a sophisticated stack of predeﬁned applications ready

for launch. When creating images with Docker, each

command executed forms a new layer above the pre-

vious one. To create a new image you use the Dock-

erﬁle, a script composed of various instructions and

arguments to automatically perform actions on a ba-

sic image. Starting from the same image it is then

possible to create multiple containers.

Kubernetes is an open source container orchestra-

tor used to automate the distribution, scaling and or-

chestration of container-based applications (Shah and

Dubaria, 2019). In Kubernetes, the pod is the funda-

mental element. It consists of a group of tightly cou-

pled containers located on the same host that share the

same IP address. Kubernetes guarantees reliability as

it can automatically restart containers that fail during

execution and terminate those that do not respond, al-

ways guaranteeing a certain number of containers in

execution.

3.2 Design and Study of Critical

Parameters

The design of the system is characterized primarily

by a study carried out in order to understand the la-

tency and reaction times necessary to avoid collisions

between vehicles, based on the examined trafﬁc sce-

narios.

Two vehicles identiﬁed by an identiﬁer number

(ID) are considered. In particular, the vehicle that

is ahead has ID = 0, while the one that follows has

ID = 1. For simplicity, in hourly laws, these IDs are

put as superscripts. Two vehicles travelling a stretch

of road at constant speed have the same deceleration

capacity ”a” and maintain a distance ”d” from each

other. It is assumed that the ﬁrst vehicle for any rea-

son is forced to brake abruptly. With t

is indicated

the sum of the latency time and the reaction time nec-

essary for the second vehicle to start braking. The

hourly law with instant started t

= 0 for the vehicle

with ID = 0 is as follows:

(

−

+ v

t t <

)

t ≥

(1)

The hourly law with instant start t

= 0 for the vehicle

with ID = 1 is as follows:

(

+ v

t t < t



+ at



t −





t ≥ t

(2)

Let suppose that due to braking vehicle 0 stopped its

run, it is possible to calculate the instant of collision

between vehicle 0 and vehicle 1 that follows, with the

assumption of consider equal the two paths traveled

by the vehicles.

)

= x



+ at



t −





(3)

Carrying out the calculations we obtain the following

second degree equation:

−



+ at



t +

+ d +

)

= 0 (4)

This equation has solutions when the delta (∆) is

greater than 0, that is:

∆ =



+ at



−4







+ d +

)



> 0

=⇒ t

(5)

Therefore, to occur a collision between the two vehi-

cles, t

must be greater than the ratio between dis-

tance and speed of the considered vehicles.

As regards the tools used in the realization, the

open source simulator Sumo (sum, 2019a) was cho-

sen to generate realistic vehicular trafﬁc in the urban

environment, so as to have data from vehicles and

to be able to process them and the Eclipse develop-

ment environment to interface with the simulator and

to be able to program it using code written in Java

language, with the library called SumoTraciConnec-

tion (sum, 2019b).

SIMULTECH 2020 - 10th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

130

To carry out the communication between vehi-

cle and MEC server and between vehicle and Cloud

server, the CoAP protocol (Puangnak et al., 2019)

was chosen using the Californium library written in

Java (cal, 2019). Each vehicle was created as a CoAP

client, and periodically makes requests to the MEC or

Cloud server, as appropriate, which acts as a CoAP

server.

While the simulator and, therefore, the various

”vehicle” clients are running on a PC, the MEC server

was hosted in a Docker container in order to have all

the necessary dependencies and to be able to run it

on another PC. In order to guarantee reliability and

greater ease of use, the Docker container for the server

was managed by the Kubernetes orchestrator.

4 SYSTEM DEVELOPMENT

The development of the system consists ﬁrst of all

in the implementation of vehicular trafﬁc using the

SUMO simulator (Lim et al., 2017). Subsequently,

the implementation of the clients, that is, the commu-

nication from vehicles to server has been developed.

Finally, the implementation of the MEC server and,

therefore, the communication from server to vehicle

has been created. The ﬁnal system is able to simulate

a trafﬁc of vehicle that communicate between them

and with the server in order to manage a collision

avoiding mechanism and, so guaranteeing security on

the vehicular environment.

4.1 Vehicular Trafﬁc with Sumo

In the realization of the system, in order to use Sumo,

a section of the road route related to the SS107 Silana-

Crotonese in Italy was initially downloaded and pro-

cessed with the command netConvert from the com-

mand line on windows, to obtain in output a xml type

ﬁle compatible with Sumo.

This ﬁle contains all the attributes relating to the

road portion. For example, there are the IDs related

to the two lanes of the roadway with the two travel

directions, the name of the considered road portion

and the maximum allowed speed expressed in m/s.

Subsequently, an XML type ﬁle was created to de-

ﬁne the vehicles travelling on the previously chosen

road portion. For each vehicle it is possible to indi-

cate its acceleration and deceleration value, id, length,

maximum speed in m/s and minimum gap with the

previous vehicle, and so on.

Finally, in order to start the simulation, it was nec-

essary to generate the XML conﬁguration ﬁle with the

.conﬁg extension. As input it takes the ﬁle related to

the road route, the ﬁle related to vehicles and an ad-

ditional ﬁle to display an alert on the map in case of

a collision between two vehicles. There is a time sec-

tion where the duration of the simulation with the step

is speciﬁed. It is also possible to output a ﬁle with the

simulation logs.

The implemented system follows a client-server

structure with CoAP communication. The Sumo sim-

ulator connected to Eclipse is used to manage client-

side communication, that is, it treats the vehicles that

are travelling in the map. This communication can be

only with a MEC server, only with a Cloud server or

characterized by a dynamic switching between MEC

Server and Cloud.

4.2 Communication Towards Edge or

Cloud

As regards the communication of the vehicle to the

server, in general, each one represents a CoAP client,

able to send information to the server through the

POST method of the Californium library (Kovatsch

et al., 2014). The information is enclosed in a mes-

sage and relates to the status of the vehicle, such as

id, speed and position. Furthermore, every command

that the vehicles receive from the server is managed to

pass it to the simulator. In order to realize this process,

a Java class has been implemented which contains the

logic related to the vehicle.

Brieﬂy, the behavior of the individual vehicle is

shown in the following ﬂow chart, ﬁgure 2.

4.3 Dynamic Switching Edge - Cloud

The implementation of the dynamic switching algo-

rithm has the aim of not overloading the Edge server

when it is not essential, deviating part of the trafﬁc

to a Cloud server. In particular, each vehicle can de-

cide whether to communicate with the Cloud server

or with the MEC server, so that the roadside device

will receive fewer packets number since part of these

will be managed by the Cloud. The decision made by

the vehicle depends on the current vehicle trafﬁc con-

ditions, with the aim of keeping the collisions number

as low as possible. Once the previous and the next ve-

hicle of a generic vehicle have been identiﬁed, this

performs a calculation in order to understand what

communication to establish, if it is better communi-

cate with the edge or cloud server. Since the instant

of collision between two vehicles depends on the total

time between latency and reaction, and the reaction,

as shown in the study done previously, for the same

amount of deceleration and speed between the vehi-

cles must be greater of the distance ratio between two

Proposal of an Automotive Collision Avoidance System based on Edge Computing

131

Figure 2: Flow chart of the behavior of a vehicle.

vehicles and the speed for collision, this ratio is con-

sidered in the calculation made by the vehicle with

respect to the following and previous vehicle in order

to choose the server. If in at least one of the two cases,

the ratio is less than the total time between the latency

of the Cloud and the reaction of the vehicle, it means

that communication to the Edge is necessary other-

wise there would be a collision choosing the Cloud.

Otherwise, communication with the Cloud is chosen.

In general, the algorithm is described by the following

ﬂow chart, ﬁgure 3.

As regards incoming communications, each vehi-

cle must check whether it has received warnings from

both the Cloud server and the MEC server. Therefore,

two CoapHandlers procedure were needed to manage

both communications asynchronously.

4.4 Server MEC

The role of the MEC server on the roadside, in the

implemented application, is to receive data from trav-

elling vehicles, process it to understand if there is a

danger or emergency braking by a vehicle, and imme-

diately notify all vehicles that follow in order to avoid

collisions. Everything is summarized in the handle-

POST method which manages every single commu-

Figure 3: Flow chart relating to dynamic switching.

nication with the client. After accepting the commu-

nication, the server decodes the received message and

performs calculations to understand if the vehicle is

braking abruptly. In particular, it makes the differ-

ence between two consecutive vehicle speeds, and if

this difference is greater than a certain threshold, it is

identiﬁed as a dangerous condition. Therefore in this

case, the server identiﬁes all the vehicles that are ar-

riving within a certain range, in order to alert them in

time and avoid collisions. The operating principle is

described brieﬂy in the following ﬁgure 4.

Figure 4: Flow chart relating to the behavior of the MEC

server.

Once the Java Server code was realized, a Docker

container could be created. This latter was then pub-

lished on a private repository created in the Docker

Hub registry so that it can be managed with Kuber-

netes. In Kubernetes a deployment has been created

where a pod characterized by the single container is

declared. Finally, a service has been created to make

the pod accessible.

SIMULTECH 2020 - 10th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

132

(a) (b)

Figure 5: Percentage of Collisions with 4G and 5G network varying vehicles distance.

(a) (b)

Figure 6: Percentage of Collisions with 4G and 5G network varying server range action.

5 PERFORMANCE ANALYSIS

To understand the performance of the implemented

system, several simulation tests were carried out.

Firstly, the collisions percentage between vehicles

varying the distance between them, using ﬁrst com-

munication with MEC Server and then with Cloud

Server, considering 4G and 5G technology was as-

sessed. In particular, both technologies have been

used from a simulation point of view, considering the

latency times that characterize each of them. The sce-

nario considered to make the measurements is as fol-

lows: four vehicles at a distance ”d” from each other,

with constant speed equal to 50 km/h, same deceler-

ation values, random reaction times between 450 ms

and 500 ms (Makridis et al., 2018) and a range of ac-

tion of the server equal to 50 m from the point where

the hazard event was generated. The ﬁrst vehicle at

a certain moment is forced to brake suddenly. The

distances ”d” expressed in meters and taken into con-

sideration are: 6, 6.5, 7, 7.5, 8, 8.5, 9, 9.5. Twenty

runs were carried out for each conﬁguration consid-

ering ﬁrst the communication with the MEC server

(4G and 5G) and then with the Cloud server (4G and

5G). The results obtained are shown in the following

graphs, see ﬁgure 5.

As it is possible to appreciate, considering the

same distance between two vehicles, the percent-

age of collisions that characterizes communication

with the Edge is lower than communication with the

Cloud. Furthermore, considering ﬁrst 4G and then

5G, in some cases the number of collisions for the

same distance between two vehicles decreases. Sub-

sequently, the percentage of collisions between vehi-

cles was evaluated, no longer changing the distance

between two, but varying the range of action of the

server with respect to the point where the hazard event

was generated. The scenario considered is as follows:

four vehicles at a distance of 7 m from each other,

with a constant speed of 50 km/h, same deceleration

values, random reaction times between 450 ms and

500 ms (Makridis et al., 2018) and a range of action of

the server equal to ”r” meters with respect to the point

where the hazard event was generated. The ﬁrst vehi-

cle at a certain moment is forced to brake suddenly.

The ranges used expressed in meters are: 15, 20, 26.

These values have been chosen respectively to ensure

that the server once only warns the vehicle following

the one that suddenly brakes, then the next two and ﬁ-

nally all three vehicles that follow. Twenty runs were

Proposal of an Automotive Collision Avoidance System based on Edge Computing

133

carried out for each conﬁguration considering ﬁrst the

communication with the MEC server (4G and 5G) and

then with the Cloud server (4G and 5G). The results

obtained are those shown in the ﬁgure 6. As you can

see, both using 4G and 5G technology, and consid-

ering communication with the Cloud server and with

the MEC server, as the range of action of the server in-

creases, the percentage of collisions decreases. This

happens because a vehicle warned in time of the dan-

ger will start braking earlier avoiding collision with

the previous vehicle. Therefore the more vehicles are

warned by the server, the lower the collision rate. An-

other parameter that has been evaluated is the percent-

age of use of the Cloud and MEC servers in dynamic

switching. This algorithm allows, especially in cer-

tain vehicular trafﬁc conditions, not to overload the

Edge and communicate with the Cloud server which

has higher performance. Three scenarios are consid-

ered in the evaluation, starting from one that is not

congested up to the congested one, as shown in the

ﬁgures 7a, 7b and 7c and the communication is made

towards cloud or edge server on the basis of latency

issues.

(a)

(b)

(c)

Figure 7: (a) Uncongested Vehicular Trafﬁc; (b) Low Traf-

ﬁc Congestion; (c) Congested Vehicular Trafﬁc.

6 CONCLUSIONS

In the paper, the attention was focused to the service

delivery paradigm called Edge Computing, realizing

a speciﬁc use case that demonstrates its importance.

In particular, an assisted guidance system for colli-

sion prediction has been developed and it has been

shown how the use of Edge technology is essential in

some circumstances where the latency parameter that

characterizes communications must be very low. This

is because, with Edge Computing, each Cloud func-

tionality moves close to the end user and therefore to

the edge of the network. Therefore in the system it

was necessary to implement an MEC server to man-

age communications with vehicles, run on a Docker

container managed by the Kubernetes orchestrator. In

order to show the improvements of Edge Comput-

ing compared to Cloud Computing, a comparison has

been made between the communication of a vehicle to

the Egde server and to the Cloud server, evaluating the

percentage of collisions of the vehicles. Obviously,

communication with the Edge allows to signiﬁcantly

reduce the number of collisions as it guarantees min-

imum latency. It has also been demonstrated from a

simulation point of view how in some cases the MEC-

5G combination is essential for this type of system.

REFERENCES

(2019). Docker: Empowering app development for devel-

opers. https://www.docker.com/.

(2019). Eclipse californium (cf) coap framework. https:

//www.eclipse.org/californium/.

(2019). Kubernetes: Production-grade container orchestra-

tion. https://kubernetes.io/.

(2019a). Sumo - simulation of urban mobility. http://sumo.

sourceforge.net/.

(2019b). Sumotraciconnection. https://sumo.dlr.de/javadoc/

traas/it/polito/appeal/traci\\/SumoTraciConnection.

html.

Ahmed, A. and Pierre, G. (2018). Docker container de-

ployment in fog computing infrastructures. In 2018

IEEE International Conference on Edge Computing

(EDGE), pages 1–8. IEEE.

Bernstein, D. (2014). Containers and cloud: From lxc

to docker to kubernetes. IEEE Cloud Computing,

1(3):81–84.

De Rango, F., Fazio, P., and Marano, S. (2008). Utility-

based predictive services for adaptive wireless net-

works with mobile hosts. IEEE Transactions on Ve-

hicular Technology, 58(3):1415–1428.

De Rango, F., Veltri, F., Fazio, P., and Marano, S. (2009).

Two-level trajectory-based routing protocol for vehic-

ular ad hoc networks in freeway and manhattan envi-

ronments. Journal of Networks, 4(9):866–880.

SIMULTECH 2020 - 10th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

134

Fazio, P., De Rango, F., Tropea, M., and Voznak, M.

(2017a). Cell permanence time and mobility anal-

ysis in infrastructure networks: Analytical/statistical

approaches and their applications. Computers & Elec-

trical Engineering, 64:506–523.

Fazio, P., Sottile, C., Santamaria, A. F., and Tropea,

M. (2013). Vehicular networking enhancement and

multi-channel routing optimization, based on multi-

objective metric and minimum spanning tree. Ad-

vances in Electrical and Electronic Engineering,

11(5):349–356.

Fazio, P., Tropea, M., De Rango, F., and Voznak, M. (2016).

Pattern prediction and passive bandwidth manage-

ment for hand-over optimization in qos cellular net-

works with vehicular mobility. IEEE Transactions on

Mobile Computing, 15(11):2809–2824.

Fazio, P., Tropea, M., and Marano, S. (2017b). Node re-

routing and congestion reduction scheme for wireless

vehicular networks. Wireless Personal Communica-

tions, 96(4):5203–5219.

Fotino, M., Gozzi, A., Cano, J.-C., Calafate, C., De Rango,

F., Manzoni, P., and Marano, S. (2007). Evaluating

energy consumption of proactive and reactive routing

protocols in a manet. In IFIP Conference on Wireless

Sensor and Actor Networks, pages 119–130. Springer.

Frnda, J., Voznak, M., Rozhon, J., and Mehic, M.

(2013). Prediction model of qos for triple play ser-

vices. In 2013 21st Telecommunications Forum Telfor

(TELFOR), pages 733–736. IEEE.

Giust, F., Sciancalepore, V., Sabella, D., Filippou, M. C.,

Mangiante, S., Featherstone, W., and Munaretto, D.

(2018). Multi-access edge computing: The driver be-

hind the wheel of 5g-connected cars. IEEE Commu-

nications Standards Magazine, 2(3):66–73.

Jain, N. and Choudhary, S. (2016). Overview of virtu-

alization in cloud computing. In 2016 Symposium

on Colossal Data Analysis and Networking (CDAN),

pages 1–4. IEEE.

Khelil, A. and Soldani, D. (2014). On the suitability

of device-to-device communications for road trafﬁc

safety. In 2014 IEEE World Forum on Internet of

Things (WF-IoT), pages 224–229. IEEE.

Kovatsch, M., Lanter, M., and Shelby, Z. (2014). Cali-

fornium: Scalable cloud services for the internet of

things with coap. In 2014 International Conference

on the Internet of Things (IOT), pages 1–6. IEEE.

Lim, K. G., Lee, C. H., Chin, R. K. Y., Yeo, K. B., and Teo,

K. T. K. (2017). Sumo enhancement for vehicular ad

hoc network (vanet) simulation. In 2017 IEEE 2nd

International Conference on Automatic Control and

Intelligent Systems (I2CACIS), pages 86–91. IEEE.

Makridis, M., Mattas, K., Borio, D., Giuliani, R., and

Ciuffo, B. (2018). Estimating reaction time in adap-

tive cruise control system. In 2018 IEEE Intelligent

Vehicles Symposium (IV), pages 1312–1317. IEEE.

Pham, Q.-V., Fang, F., Ha, V. N., Le, M., Ding, Z., Le,

L. B., and Hwang, W.-J. (2019). A survey of multi-

access edge computing in 5g and beyond: Funda-

mentals, technology integration, and state-of-the-art.

arXiv preprint arXiv:1906.08452.

Porambage, P., Okwuibe, J., Liyanage, M., Ylianttila, M.,

and Taleb, T. (2018). Survey on multi-access edge

computing for internet of things realization. IEEE

Communications Surveys & Tutorials, 20(4):2961–

2991.

Puangnak, K., Puisamlee, W., Puangnak, K., and Rach-

siriwatcharabul, N. (2019). Evaluation of mqtt and

coap for vehicle trafﬁc monitoring. In 2019 16th

International Conference on Electrical Engineer-

ing/Electronics, Computer, Telecommunications and

Information Technology (ECTI-CON), pages 915–

918. IEEE.

Santamaria, A. F., Fazio, P., Raimondo, P., Tropea, M.,

and De Rango, F. (2019). A new distributed predic-

tive congestion aware re-routing algorithm for co 2

emissions reduction. IEEE Transactions on Vehicular

Technology, 68(5):4419–4433.

Santamaria, A. F., Tropea, M., Fazio, P., and De Rango,

F. (2018). Managing emergency situations in vanet

through heterogeneous technologies cooperation. Sen-

sors, 18(5):1461.

Shah, J. and Dubaria, D. (2019). Building modern clouds:

using docker, kubernetes & google cloud platform. In

2019 IEEE 9th Annual Computing and Communica-

tion Workshop and Conference (CCWC), pages 0184–

0189. IEEE.

Sharma, S. et al. (2019). Vehicular ad-hoc network: An

overview. In 2019 International Conference on Com-

puting, Communication, and Intelligent Systems (IC-

CCIS), pages 131–134. IEEE.

Socievole, A., De Rango, F., and Coscarella, C. (2011).

Routing approaches and performance evaluation in

delay tolerant networks. In 2011 Wireless Telecom-

munications Symposium (WTS), pages 1–6. IEEE.

Song, W. and Jiang, B. (2015). Visualization of iov warning

data based on warning velocity association model. In

2015 8th International Symposium on Computational

Intelligence and Design (ISCID), volume 2, pages 30–

33. IEEE.

Xiong, K., Leng, S., He, J., Wu, F., and Wang, Q. (2019).

Recouping efﬁcient safety distance in iov-enhanced

transportation systems. In ICC 2019-2019 IEEE In-

ternational Conference on Communications (ICC),

pages 1–6. IEEE.

Yang, F., Wang, S., Li, J., Liu, Z., and Sun, Q. (2014). An

overview of internet of vehicles. China communica-

tions, 11(10):1–15.

Zhou, B., Lee, Y.-Z., Gerla, M., and De Rango, F. (2006).

Geo-lanmar: a scalable routing protocol for ad hoc

networks with group motion. Wireless Communica-

tions and Mobile Computing, 6(7):989–1002.

Proposal of an Automotive Collision Avoidance System based on Edge Computing

135