A Review On Smart Parking Systems In Metropolitan

Cities Using IoT

Ankit Patel Meghan Patil

Department of computer science Department of computer science,
, B Tech Integrated B Tech Integrated Mihir Prajapati
NMIMS UNIVERSITY, MPSTME NMIMS UNIVERSITY, MPSTME Department of computer science,
Mumbai, India Mumbai, India B Tech Integrated
[email protected] [email protected] NMIMS UNIVERSITY, MPSTME
Mumbai, India
[email protected]

Abstract— In today’s world Internet has become an One of the major problems faced by humans in this decade is
integral and important part our daily life. Most of the finding an empty parking space. Researchers have found out
things we do are connected to the internet is some way or that in INDIA, there are about 50,000 vehicles registered
the other. IoT platform enables us to implement wide every day. According to a recent survey conducted by Times
range of applications. It makes use of ultrasonic sensors, of India, it was found that on an average 685 cars are being
controllers, and interpreters to make to make the overall registered on daily basis. Mumbai had nearly 32 lakh vehicles
system smart. One such application is parking system. last November (Fig 1). So we can imagine the rate at which
Smart parking system can be implemented by the use of traffic and the number of vehicles is growing.
IoT technology which provides various services such as
information about parking spaces, the real time
availability and unavailability of slots. Increasing traffic
and number of vehicles, results in mismanagement of
parking spaces. One has to spend a lot of time to find a
vacant slot. IoT technology allows us to overcome such
issues by using various components which helps to make
the parking system efficient. In this paper, we have
reviewed various systems that have been proposed and
analyzed each of them and solved some of the issues that
were encountered in these papers.

Keywords— IoT, Smart Parking System, Smart City

I. INTRODUCTION Fig 1: Number of registered vehicles in India

Internet of Things is a concept where physical objects are
connected to the internet. The term internet refers to scenarios Most of the provided parking systems are considered to be just
where network connectivity and real-time computing comes Informative Parking System (IPS) rather than “a smart parking
into the picture. It deals with physical objects like vehicles, system”[1]. One has to spend a lot of time to find a parking
smart thermostat systems heart monitoring implants, Wi-Fi area. And even after successfully locating one, he/she faces a
enabled washers/dryers and other objects embedded with problem to reach that area. Even after reaching the parking
software and sensors. The concept of combining computers area the person takes a lot of time to locate the exact spot. In
and sensors has existed for decades but the recent market most of existing systems, the driver has to spend a lot of time
trends has attracted the Internet of things closer to widespread driving around the parking area to find an empty slot which
reality. As the arena of Internet of Things is growing, the results in wastage of many valuable resources such as time,
applications based on it are also growing. In this paper, we fuel and so on. The result of a survey conducted recently
proposed an IOT based smart online parking system which showed that in India, it takes about 8.2 minutes for driver to
provides an easy and comfortable way of finding a parking lot successfully park their vehicle (low traffic). However, this
and saves time and money. value may vary depending upon the density of traffic. In cities
like Mumbai and Delhi, it may take around 20 min. The
majority of time that is consumed in the parking process is II. LITERATURE REVIEW
caused by:
(A) at the entry point where the driver has to get the parking A. Related Works
card. In the previous works, several solutions were proposed in [2],
(B) time taken to locate an empty space. [3], [4], [5], [6] for improving the existing parking system.
In the previously proposed papers for Smart Parking System, Each of the proposed method is summarized is discussed
the problem that was faced was that, once a user books a slot below:
through the mobile application, the details of the slot and level
number is provided to the user through the application. But, if [2] In this paper the Author has proposed a smart parking
accidently the user parks in some wrong location, then the system which makes use of RFID (Radio Frequency
entire system will be in error. Identification) and GSM (Global System for Mobile)
In this paper, we propose a solution to the problems mentioned technology which intends on providing a smart solution to
above by introducing an IOT based smart parking system deal with vehicle thefts. The main purpose of this technology
which makes use of a mobile application to provide various is to make use of this wireless technology along with an
smart services such as application to solve the security issues that are encountered in
 Locating the available parking area. an organization. The Author focuses mainly on the security
 Real time slot vacancy detection. aspects rather than focusing on the way the vehicles must be
parked.
 Navigation to a specific parking area. Working of RFID AND GSM
 Booking a slot (online/offline). The main components of the RFID system are:
 Detecting the car number plate 1. RFID interrogators
2. RFID labels
 Recording the location of the parked car.
3. GSM kits
 Calculating the fare and payment.
The GSM kit is used to send SMS to an authorized user. RFID
The proposed system also provides an efficient way to ensure Readers and Tags are used to control the check in and check
that the driver parks the vehicle in the allotted slot only. outs.
A. ISSUES RFID system consists of 3 components in combination of two
 Transceiver (Transmitter and receiver) and antenna
1. In our proposed system i.e, A Smart Parking System
which forms RFID READER.
Using IOT, the problem that was faced was that, once
 Transponder (Transmitter and responder) and antenna
a user books a slot through the mobile application,
which forms RFID TAG.
the details of the slot and floor number are provided
to the user through the application. But, if accidently When the RFID READER emits a radio signal, it activates the
the user parks in some wrong location, then the entire transponder and the RFID TAG is read by the READER.
system will be in error. There are two types of transponders, which correlate to the
2. One of the most common problems faced by the two major types of RFID tags Active transponder contains a
drivers is that they are unable to find a vacant slot battery that periodically transmits signals. Passive
when they arrive at parking area. This problem may transponders do not have their own energy source and they
rely on the energy given off by the reader [2].
be caused due to various reasons like, peak hour
Advantages
parking etc. i.e, how the amount of time can be  High security.
reduced for finding a vacant slot and efficiently park  Eliminates the chances of human error.
the vehicle?  Reduces operational cost.
3. What if multiple users are accessing the same slot at
the same time during online reservation? Disadvantages
4. What if the user wants to park their vehicle only for a  If the RFID reader fails, the entire system will fail.
certain period of time?  Check-in and check-out is only allowed in case of
5. How can we reduce the amount of energy consumed registered users.
by the cameras installed at each parking slot?
6. What if the ultrasonic sensor fails to detect a vehicle [3] Author has proposed a smart parking system model which
present at the slot or detects the vehicle outside the made use of an ultrasonic sensor which was fit in the ceiling
slot area? of each car. Ultrasonic sensors are operated on the basis of
echolocation. It emits a sound wave which hits the vehicle and
reflects back to the sensor. The sensor is connected to a server
which collects all the data and updates it in the application’s
LEDs which are used to indicate whether the slot is empty or obtained from the sensor is forwarded by the
not. Red color indicates that the slot is occupied and green Raspberry Pi and updated in the mobile application.
color indicates that the slot is free to be utilized. The heart of  Camera- instead of asking the user to manually enter
this system is the embedded controller which interacts with the the license plate number into the application. The
Raspberry pi and the Camera Module.
cameras can be installed which can extract the text
Step wise representation; from the number plate and store it in the database.
This also provides evidence that a particular vehicle
1. Driver uses the mobile application to locate was present here at a particular time. Thus improving
the nearest available parking area. the security of the parking area.
The driver selects an appropriate option from all the listed  Raspberry Pi3- it controls the hardware data, which
parking areas enables the sensor and camera to work together. It
provides a connection to the cloud database.
 Mobile application will send the requested  Mobile Application- it is software that is installed on
information to cloud. It responds to the request and the users Smartphone which enables them to access
provides a map like structure of the nearest available various services from the application.
parking areas.  Cloud- used to store the data retrieved from the
 The available/unavailable slots are displayed in the users and the sensors. It passes the information to and
application using graphics, colours and symbols from the user and the sensor respectively. It is also
along with the cost. useful to store the details of all the booking and
payments.
2. The driver is provided with the navigation to
the selected parking area.
 The driver follows the route specified by the
application to get to the parking area.
3. The driver books the nearest available slot
and makes the payment.
 After the slot is booked the user selects the amount of
time they wish to park the car. According to the
selected time, bill is generated and online payment is
made by the user.
4. Driver is guided to the slot and the location of
the parked car is recorded.
 The driver is navigated to their respective slots and
after the car is parked, the notification to fill the
license plate number is indicated on the application.
 Once the information is filled, the current location of
the vehicle is noted and saved for later use. Fig 2: Flowchart of various elements architecture
 The driver is provided with an e-bill which can be
accessed from the application.
5. The driver is notified 15 minutes prior to the
slot time expiration. The driver then returns
to the car and exits the parking area.

A notification is sent on the application along with location of

the parked car 15 minutes prior to the timer expiration. If the
user does not checkout within the allotted time, extra charges
will be applicable. And finally, the driver tracks the vehicle
position and returns to the location and leaves the area. To
implement this proposed model, various components have
been used. The system architecture makes use of following Fig 3: Ultrasonic sensor detection and LED color updation.
technologies and techniques:
 Ultrasonic Sensors- it is used at each slot to detect
the presence or absence of a vehicle. The data
  Fear of breakdown- there is always a fear that some
hardware system may breakdown due to some
reasons. Therefore, frequent maintenance is must.

[4] Author has proposed a system that serves as an e-valet

and helps to driver/customer to find the empty parking slot in
the parking area. The proposed system makes use of a NFC
technology (near field communication).
Fig 4: Camera to record the number plate text. In the proposed system the Author has introduced an e-Valet
parking service which is similar to the original Valet systems.
When the driver brings the car close enough, the sensor is turned on The main purpose of this system is to have an employee or a
and the image of the license plate is taken and the data is forwarded third person who takes up the task of finding the empty slots
to the cloud. in the park area. Similarly, e-Valet is concept of using
electronic means to remove the need of a third person but still
Advantages/Disadvantage functions like valet service [1].
The e-Valet service provides an easy and efficient way for the
Advantages customer to look and book the available parking slots in the
park area. It also enables pre-booking i.e, booking the slots in
 High parking efficiency- the users can find the best advance through a mobile based application.
spots available with the least efforts. Using iSCAPS, the e-Valet mobile application, users can make
 Saves time- the user no longer has to spend a long a request for reserving a slot. Before allocating the slots, the
amount of time searching for an empty slot. system will look into the database to check the availability of
slot. The slots which are vacant can be reserved by the
 Enhanced security- since all the user information is
customer.
saved in the database, in case of any accident the Customers who wish to enter the parking area make use of
stored data can be used to find the details of the their smart phones or rent out the NFC cards. The user has to
owner. tap the NFC enabled device on a NFC reader which will
 Online payment system- unlike the traditional display the available number of parking slots at each level on
offline payment system which provides a printed the LCD screen. An empty slot will be assigned to the user and
copy (paper), here we can make the payment online the LED at that lot changes from green to blue i.e, it is waiting
and obtain an e-doc of the details of the payment for the customer to park the vehicle. The same procedure is
performed when the customer wants to exit the parking area.
which is eco friendly as it prevents unnecessary
Additionally, the amount gets deducted from the NFC device.
usage of paper. The system also provides a solution to the problem when the
user forgets his/her parked location. To determine the location
of the parked vehicle, the user only needs tap the NFC device
at an NFC kiosk machine which provides the location of the
vehicle as well as the available balance.
Disadvantages Table 1: Components and their working
 Cost of implementation is high- using several
hardware and software devices increases the overall
cost.
 Replacement of hardware is difficult- once
installed, it is difficult to modify the sensors and
cameras and also the cost of replacement is also high.
 Failure of the Cloud- if the cloud database fails the
entire mobile application will break down as it’s the
core part of the entire application.
 It may be confusing for unfamiliar users- the users
may find it difficult to use the application of follow
the instructions specified by the navigation system.
 Maintenance is high- the hardware needs to be
checked frequently to ensure that it is working Advantages
properly.  More convenient
  Eco friendly Author has proposed a framework using Mobile Cloud-
 Saves time Computing capability and emphasizes cloud-assisted
architecture for parking reservation system.
Traditional parking garages:
Disadvantages
The context information of each parking space detected by a
 Can be costly Wireless Sensor Networks(WSN), and the internet. The data
 Complex collected by the sensors are moved to the cloud storages and
then selectively transmitted to the users. Also, the status of the
parking garages may be dynamically published on nearby
[5] In this paper, we propose Vehicular Cyber-Physical billboards to users.
Systems (VCPS) along with the cloud support. Here, each Dynamic parking services:
layer may provide multiple context-aware services. In this scenario, we consider a situation in which we may
CLOUD-ASSISTED CONTEXT-AWARE temporarily park a vehicle along the road. For example, during
ARCHITECTURE rush hours, there is usually heavy traffic. Therefore,
In this section, author has proposed a context-aware considering this information such as road conditions and rush
architecture with mobile cloud support and two crucial service hours, we may dynamically arrange parking services for a
components. Fig 5 shows a cloud-assisted context-aware very wide road. With the support of many new technology as
architecture. Mobile Cloud-Computing (MCC) and Wireless Sensor
MULTI-LAYERED ARCHITECTURE Networks, the traffic authorities can carry out the dynamic
In the vehicle range, onboard equipment (OBE) installed in management of this kind of service.
the vehicle can provide all kinds of services (e.g., GPS In the proposed framework, analysis of three aspects,
Navigation, Entertainment). In the location computational including service planning of traffic authorities, reservation
layer, roadside equipment (RSE) deployed at strategic service process and context-aware optimization is being done.
locations can exchange information with OBE installed on Decision making of traffic authorities:
vehicles passing by. Vehicles outside the range of RSE may The decision making of the proposed scheme depends on
still be connected to the infrastructure network via many factors, such as historical traffic flow capacity, road
neighboring vehicles. This infrastructure network can generate conditions, weather conditions and traffic flow forecasting. In
accurate real-time traffic information. order make effective predictions, researchers need to do data
In Fig-5, an inter-cloud environment in CVC includes multiple mining and discover useful information and knowledge from
cloud systems running with different policies and different collected big data.
environment with each other to share resources so that end-to- Parking reservation services:
end QoS can be maintained even in the event of large As shown in Fig-6, the status of the parking space can be
fluctuations in computing load which cannot handle by a monitored as shown in the corresponding system, and
single cloud system.[9] concurrently updated in the traffic cloud. Through smart
phones, users can quickly obtain parking space’s information.
Within a given time, we may log into the traffic cloud and
subscribe to a parking space.

Fig 5: example of a Context-Aware architecture

WORKING
 Figure 6: Context aware dynamic parking services. information of parking space. It is an improvement of parking
management part of existing ITS infrastructure which can
Context-aware optimization solve the problems mentioned above efficiently.
The context information includes not only road conditions and System structure
the status of the parking spaces, but also the expected duration A) Sensor Networks on the street
of a parking as well. The purpose of the proposed system is to
optimize the best locations for drivers. For the parked To implement this system, we should deploy sensor networks
vehicles, the expected duration of parking can be uploaded to on the street where parking management is needed.
the traffic cloud and the information is shared with the drivers Deployment method can be seen in Fig. 7. It does not need to
simultaneously. In this way, even when the parking garages embed sensors on the road deeply which means low cost.
has no empty slots, drivers still can inquire and get the desired Every sensor detects the presence of vehicle by measuring
services by context-aware optimization. disturbances in the Earth's magnetic field. In addition, each
The proposed context-aware dynamic parking service offers sensor has luminous device, such as LED, which indicates the
the solutions to overcome the parking problems faced by the reservation information of the parking space.
people in their daily life and improving the quality of Context-
aware VCPS. Many technologies such as WSN, traffic clouds,
data mining and big data are enabling application scenario to
become reality.

[10] But, due to the shortage of available land per person and
city planning being lagging behind, the problem of parking
difficulty becomes increasingly severe. To reinforce the
management of roadside parking, to call for reasonable
roadside parking and to develop a novel roadside parking Fig 7: System Schematics
management system become an effective means to solve the B. Other Parts of the System
existing problems. Small control centre (i.e., base station), which is located at the
Roadside Parking System and Sensor Networks intersections, is responsible for collecting data from the
Chronologically, the development of roadside parking nearest sensor and judges the state of parking space (i.e.,
management system mainly goes through several stages as vacancy or occupancy). Each small control centre connects to
follows: main control centre through wired network according to mesh
1) Stage of Charging by Manpower: This type of system network topology and architecture. The parking space state
needs to delimit the parking areas and missionaries are information collected by small control centre will be
required to charge and find parking space for drivers. Thus, transmitted to main control centre finally. Main control centre
this system has low efficiency. Although it means a low conducts classification process so that users can query it.
expenditure at first, the cost of manpower is high.
2) Stage of Charging by Parking Meter: Parking meter
system is still the basic element of infrastructure of many 4) System Design
traffic management systems. It requires drivers to find parking A. System Overall Design:
space themselves, but it charges by parking meters. Its In this system, there is a sensor at every parking lot and
advantage is that it leaves out the cost of manpower, but, at the sensor is responsible for detecting the presence of vehicles.
mean time, there are many disadvantages. For example, the Every sensor can communicate with the micro-operating
drivers have to find parking space blindly without any system embedded in the vehicle to identify the information of
navigation information that wastes a lot of time. Furthermore, vehicle. The system detects the presence of vehicles by
installation of parking meters needs a lot of capital devotion measuring the disturbance in the Earth's magnetic field. To use
and affects the city's appearance too. In particular, it is hard to power efficiently, microcontroller in each sensor can switch
solve the problem of high power waste. itself to a dormant state. Microcontroller can wake it up
3) Stage of ITS: Currently, the Intelligent Transportation periodically so that sensor can collect the information of
System (ITS) has been put into run. ITS requires video parking space. As sensor transmits data in a short distance,
cameras to be installed and missionaries watch the picture every sensor communicates with sensors nearby, but several
formed by video camera in control room which is located at sensors can form sensor networks. The sensor networks in the
remote end and judge the condition of parking space according system degrade to a linear network which is beneficial to
to the picture. Its disadvantage is that it can't run reliably when avoiding system choke point of data transmission. Every
the weather is awful, such as rain, fog and snow. Although sensor transmits the collected data to small control centre at
some systems use radar, sonar or laser radar instead of video the intersections by passing through several sensors. In
camera, which enhance the reliability of the system, the addition, small control centre is responsible for counting users'
system costs more and has high power waste. This paper parking time and transmitting parking time and parking space
presents a sort of system which uses sensor networks to collect state information to main control centre.
 configure the network information. The partial simulation of
B. Node structure the system is illustrated in Fig. 6:
Fig. 8 describes data flow direction of sensor node, from
collecting data to transmitting to neighboring node. Firstly,
magnetic sensor detects the presence of vehicle by measuring
the disturbances in the Earth's magnetic field. The collected
signal is signified and sent to microcontroller and then
microcontroller sends the signal to radio transmitter after
relevant judgment. Finally, radio transmitter transmits the
signal to neighboring node by antenna. The power of the
system is supplied by a battery Fig. 9 introduces the integral
structure of sensor networks and routing information of
parking space state. Several nodes can compose a cluster and
data is transmitted from originator to sink node and then to
other networks, such as Internet.

Through certain measurement. The power waste of functional

units in sensor node is listed in the TABLE 2 below.

Table 2: Power waste table

Fig 8: Sensor Node Structure Schematics

When sensor node is not in an active mode, its power waste is

Fig only about 130 ns-A and when in an active mode, its power
Fig 9: SensorNetwork Structure waste is dependent on the data size transmitted by sensor
System Simulation: nodes. Since the data size is small, the power of node is low.
To make the development of the system more convenient and
fast, we use NS2(network simulator 2) to simulate network III. RESULTS/INFERENCES
part of the system.
A. Solutions For The Issues
A Brief Introduction to NS2:
NS2 is written in C++. For the sake of making it convenient to 1. To overcome this problem, we have introduced
use, its front end is controlled by OTcl language (object- RFID technology which consists of two main
oriented TCL scripting language). As NS2 supports the components RFID tag and RFID reader. Firstly,
simulation of sensor networks, it can simulate the system when the user books a slot, the entire details of
effectively. In addition, NS2 has a visual windows-based that information including the user’s information
program called Nam (Network Animator) which can show the
is stored in the database. Once the user enters the
result of simulation. To use Nam to simulate, we should create
a trace file first. In the trace file, there is topological Parking Area, the details of the slot location such
information, such as node location, establishment of data link as level number, and slot number will surely be
and time of transmitting data packets etc., which is necessary provided. When the user attempts to park into a
to simulate the system. NS2 requires writing Tel scripting to slot, the RFID tag which is embedded within the
 QR code will be read and if the details are algorithm such as FCFS (first come first serve).
matched then the user can park his/her car. Else So the customer who initiates the payment first
the user will be notified about the same. Hence, will be allowed to book the slot and the other
ensuring that no user can park their vehicle at a person will be notified about the same and will
wrong slot. be redirected to the next available slot.
2. To overcome this problem, we have made use of 4. In such cases when the user only wants to park
a NFC technology which works as E-Valet and the vehicle for a specific period of time, the
determines the number of vacant slots on each system called parking meter can be used. In this
level. Customers who wish to enter the parking system, once the presence of a vehicle is
area make use of their smart phones or rent out detected by sensors, the led changes from green
the NFC cards. The user has to tap the NFC to red. As soon as the led changes to red the
enabled device on a NFC reader which will timer starts and as soon as the vehicle leaves the
display the available number of parking slots at led changes to green and the timer stops. Once
each level on the LCD screen. We can integrate the timer stops, the cost will be calculated
NFC technology along with Dijkstras algorithm accordingly.
to allocate the closest parking slot hence
reducing the amount of time taken to park the 5. To overcome this problem, we can make sure
vehicle. In case of parking area having multiple that the camera is only turned on when required.
entry points, the empty slot which is closest to This can be achieved by using the data generated
the customer is determined by using dijkstra by the sensors and as soon as the led turns red
algorithm and the slot closest to the entry point is i.e, a vehicle is present, the camera will be turned
allocated. on.
3. In order to eliminate the problem of concurrent
access by multiple users, we can use scheduling
B. Comparison Table
Table 3: Comparison of various Methods

Parameters Ultrasonic RFID NFC Infrared ITS RSE

Sensors sensors
Cost Low Moderate Moderate Low High High
Accuracy Moderate High High Moderate High Moderate
Range 50-70m 100m Few cm 150cm - -
Efficiency Moderate High Moderate Low Moderate High
Maintainability High High Low Moderate Moderate Moderate
wastage respectively. Also, after the advancements in
current Image processing
IV. CONCLUSION
techniques, we can eliminate the use of multiple
The growth of Internet of Things (IoT) and Cloud cameras of detection by using a wide angle camera
Computing has led to advancement in parking which when integrated with image processing and
technology which enables users, the easy and neural network techniques can increase the efficiency as well
efficient way of booking parking slots. This real-time as reduce the overall cost of the system. Also the roadside
system provides availability of unreserved parking parking facilities can be improved by the use of techniques
slots in a particular area. This paper is intended to such as context aware parking services which makes use
improve the parking facilities of a city. various OBEs (On Board Equipment) and RSEs (Road Side
The use of various sensors such as ultrasonic and Equipment) to improve the overall system performance.
infrared sensors allows for tackling and monitoring of
data about each slot in real time. Various technologies
such as RFID and NFC have to use to provide
enhancements in security and reducing the time
 [6] Chen Wenzhi and Liu Bai, "A smart roadside parking navigation system
References based on sensor networks for ITS," 2006 IET International Conference
on Wireless, Mobile and Multimedia Networks, hangzhou, China, 2006,
[1] R. Lookmuang, K. Nambut and S. Usanavasin, "Smart parking using pp. 1-4.
IoT technology," 2018 5th International Conference on Business and [7] S. Shinde, A. Patil, S. Chavan, S. Deshmukh and S. Ingleshwar, "IoT
Industrial Research (ICBIR), Bangkok, 2018, pp. 1-6. based parking system using Google," 2017 International Conference on
[2] L. Kumar, M. H. Khan and M. S. Umar, "Smart parking system using I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC),
RFID and GSM technology," 2017 International Conference on Palladam, 2017, pp. 634-636.
Multimedia, Signal Processing and Communication Technologies [8] B. M. Mahendra, S. Sonoli, N. Bhat, Raju and T. Raghu, "IoT based
(IMPACT), Aligarh, 2017, pp. 180-184. sensor enabled smart car parking for advanced driver assistance system,"
[3] R. Lookmuang, K. Nambut and S. Usanavasin, "Smart parking using 2017 2nd IEEE International Conference on Recent Trends in
IoT technology," 2018 5th International Conference on Business and Electronics, Information & Communication Technology (RTEICT),
Industrial Research (ICBIR), Bangkok, 2018, pp. 1-6 Bangalore, 2017, pp. 2188-2193.
[4] J. T. Ang, S. W. Chin, J. H. Chin, Z. X. Choo and Y. M. Chang, "iSCAPS [9] F. Toutain, A. Bouabdallah, R. Zemek and C. Daloz, "Interpersonal
- Innovative Smart Car Park System integrated with NFC technology context-aware communication services," in IEEE Communications
and e-Valet function," 2013 World Congress on Computer and Magazine, vol. 49, no. 1, pp. 68-74, January 2011.
Information Technology (WCCIT), Sousse, 2013, pp. 1-6 [10] S. Al-Sultan, A. H. Al-Bayatti and H. Zedan, "Context-Aware Driver
[5] J. Wan, D. Zhang, S. Zhao, L. T. Yang and J. Lloret, "Context-aware Behavior Detection System in Intelligent Transportation Systems," in
vehicular cyber-physical systems with cloud support: architecture, IEEE Transactions on Vehicular Technology, vol. 62, no. 9, pp. 4264-
challenges, and solutions," in IEEE Communications Magazine, vol. 52, 4275,Nov2013.
no. 8, pp. 106-113, Aug. 2014.M2