Turkish Journal of Computer and Mathematics Education Vol.12 No.

9 (2021), 1321-1330
Research Article
A Role of an Edge Computing Technologies for the Internet of Things in Smart Cities

*R.Deepa1, M.Krishnamoorthi2
1
Research Scholar, Department of Computer Science and Engineering, Dr.N.G.P Institute of Technology,
Coimbatore, Tamilnadu, India
2
Associate Professor, Department of Computer Science and Engineering, Dr.N.G.P Institute of Technology,
Coimbatore, Tamilnadu, India
1
[email protected]

Article History: Received: 10 January 2021; Revised: 12 February 2021; Accepted: 27 March 2021; Published
online: 20 April 2021

Abstract: Data-thorough analysis is the biggest contest in smart cities, due to the ubiquity of different types of devices, with
the number of device explosions and the numeral of Internet of Things (IoT) rest area attached to smart cities. Using shrewd
computing technology to create critical infrastructure and services for the city - including city administration, education, health,
public safety, land, shipping and services - is smarter, more connected and efficient. In this paper, we explore the important
technologies and structures already built in the literature to identify the right people for inclusion in smart cities for computing
development.

Keywords: Internet of things, edge computing, smart city, fog, cloud.

1. Introduction

Implausible improvements in the daily consumption of automated infrastructure and software were leading to
significant moves forward in communication networking and the introduction of the Internet of Things (IoT)
concept. IoT is an evolving example of communication, in which devices can act as objects or "things" that can
sense their environment, communicate, and share data over the Internet. By 2022, the IoT network will connect
one trillion IP addresses or artifacts to the Internet. Recently, the IoT prototype used to create smart atmospheres
with different application domains and associated services, also including smart cities and intelligent homes. The
objective of creating such that the intelligent environment can make human life more constructive and convenient,
addressing the living environment, carbon emissions and essential goods. The objective is directly expressed in
the significant development across distinct networks of accessible IoT based services and apps. The Internet of
Things is a physical device smart connectivity that drives huge gains in effectiveness, company development, and
quality of life [4]. In this paper, the Internet of Things solicitations and contests explained in detail. Section II
describes the literature review of Smart Cities Architecture. Section III gives IoT edge computing, fog computing,
and cloud computing. Section III gives edge-computing models in brief. Section IV is about tools availability, and
finally, section V presents the conclusion about this work.

2. Internet of Things

Internet of Things described as an environment where sensors and actuators interconnect and connect to the
Internet. All objects related in the IoT system speak in the same language, understand, and acts without any human
intervention. Originally, the word Internet of Things discovered in the presentation tile describing the Kevin
Ashton. This explains about sensor tagging for products of large consumer company to make the supply chain
management robust in 1999. Even the Internet of Things named very late, the concept of the connected system is
not new, and there were many similar developments in progress. The coco-cola machine was connected to the
internet during 1980 and monitored for availability, reducing unnecessary trips to the distributor. Sethi et al. [1],
has reviewed the basic IoT system, Architecture, protocols available, and applications of IoT, and proposed
taxonomies for further IoT are shown in the following figure 1.Tiwary A et al. [2], Tiwary has identified IoT
features like Artificial Intelligence enablers, creates more linked systems, and an active interaction network, and
also describes customer engagement, technology management, waste reduction and increased data collection are
becoming more and more evident in the form of IoT features. In addition, security, privacy, complexity, flexibility
and compliance are significant disadvantages of the Internet of Things.

1321
 R.Deepa, M.Krishnamoorthi

Figure 1: Taxonomy of Research in IoT Technologies

Sharma V et al. [3], has stated better security and surveillance, disability assistance, automated industrial
control, smart agriculture, home, and building automation are the major applications of Internet of things, but this
application of IoT is not limited, and currently smart city umbrella of all IoT applications and emerged as a critical
field of study are shown in the following figure 2. Sarhan, Q.I. [4] had a detailed review of the experiments and
focuses of the Internet of things. He has categorized the challenges in Global use, Standards and Protocols,
Operational and Technical, and Data and Software. He stated, as privacy and security were the biggest challenges
when IoT enabled for global use. Data integrity, Data Authentication, Data Confidentiality, and Trust and
Governance affects the global use of the Internet of things. In addition, he detailed how interoperability, mobility,
scaling, human-machine interaction, and dependability are the major operational and technical challenges of IoT
are shown in the following figure (Fig.3).

Figure 2: Applications of Internet of Things

1322
 A Role of an Edge Computing Technologies for the Internet of Things in Smart Cities

3. Literature Review

Figure 3: Challenges in Internet of Things

Gianni Pasolini et al. [5], presents about the smart city will be smarter when there is less or nil natural resource
wastage, better transport, environmental carrying, better emergency supporting and, recently, the smart city
concept emerged as a groundbreaking solution to solving the problems facing new urban areas. Rapid spread, the
aging of buildings, the movement of vehicles, and the provision of energy, personal security and Data protection
are some of the problems that have a major effect on the quality of life and health of people, and on the economic
and environmental sustainability of human operations. This factor is certainly important because, by 2050 people
living in urban areas are predicted to increase. IBM terms a clever city like "a city that makes use of all the mutual
knowledge today to better understand and monitor its operations and maximize the use of limited resources.

3.1 Existing Research

3.1.1 Architecture
Initially, Jiong jin et al. [7], was one of the first researchers describes due to increasing population of a city The
city challenges the challenges of using limited resources more efficiently meeting the needs of city dwellers,
workers and visitors, and improve the length of smart cities' critical IoT building blocks, and their communications
and policies and resolutions. Moreover, calculate the requirements and start the noise mapping work in conjunction
with IoT City. Roozbeh Jalali et al. [8] had a detailed review of the city's population presents significant
experiments, containing air toxic waste, movement crowding, health problems, vitality and discarded management.
Solutions to these threats include the coordination of multiple ICTs into urban carvings. Aseel Alkhelaiwi et al.
[9], introduce the aggregate of data to be shown to the cloud is very important and there are bandwidth, network
and storage challenges. It provides a data reduction service and proposes lossless compression steps for congestion,
accelerometer readings and single-precision floating-point data obtained from GPS coordinates. It proves floating-
point compression to reduce the cost of transmitting large amounts of data. Federico Montori et al. [10], gives the
architecture Sensquare can handle different data cradles from an exposed IoT stand and crowd sourcing battles,
and demonstrate integrated entrance for customers. Paula Ta-Shma et al. [11] presented analysis of comprehensive
historical data sets and seemly for real-time analysis, it is complex, to solve this issue proposed the hut architecture
it uses open source components to optimize for Big Data applications. Jianhua He et al. [12] Introduces a new fog-
computing model with specialized computing infrastructure and simple, functional modules that can reduce
potential problems with slow cloud computing response. It has evolved to evaluate the FOG-based analytics service
and the proposals for QoS management. Experimental findings show the analytics services' potential for
influencing multitier fog and planned QoS schemes. Bo Tang et al. [13], Fog extends computing to the edge of
computing networks, to rectify the challenge of data-intensive analysis in the smart cities. It introduces a Classified
Distributed Fog Figuring Design for Big Data Enquiry in Smart Cities. Computational and storage networks require
latency-sensitive applications at each layer as well as provide a fast control loop to help make sure data security
elements are efficient.

P.K Sharma et al. [14], Represented the embedded system in the central network and at the edge servers. This
relies on three technological advancements: Computing for Fog, SDN, and Block chain. This architecture supports
full functionality, data package in proper time, high interoperability, preservation, explosiveness and low power
consumption. It eliminates the end-to - end transmission delay between IoT systems, machine resources and
existing network lateral load as compare to direct communication protocols. Moreover, loading files in the cloud

1323
 R.Deepa, M.Krishnamoorthi

is a more appropriate path. Pradeep Kumar Sharma and others. [15], with the potential for developing software-
defined networking and block chain technology, Novel network architectures for smart cities reported. To achieve
productivity and address existing limitations, Smart City Networks divided into two classes, Argon 2, for core
network and edge networks, as well as securely distributed smart city networks using the block chain method.
Implementation of work based (energy) scheme. Zaheer Khan et al. [16], has described sensible cities consumption
a range of ICT solutions to touch upon tangible-world inner-city challenges. It includes environmental property,
socioeconomic innovation, and democratic governance, higher public services, designing, and cooperative
decision-making, and Converts application-specific knowledge into an assistive application of information and
knowledge to facilitate urban formation and policymaking. Since the ICT angle, the prospect of identifying capitals
is aided by smart hardware and software packages, e.g., IoT, RFID, sensible phones, detector nets, sensible house
utilizations, and capability to accomplish and method giant-weighbridge knowledge exploitation cloud computing
while not compromising data security and voters privacy. Zaheer Khan and others. [17] Through ICT, it is
challenging to understand the smart cities that process integration data from a variety of sources and support the
information provided by the services used by voters and public administration. Integrating large amounts of data,
collecting data in multiple formats, correlating such information to the right issues and events, analyzing records
to reduce expedient figures and visualization, and the perpetually increasing number of such Information
Management. Solving these defies requires a multi-pronged slant that includes information architecture
standardization, information integration processes, process processes and storage infrastructure. In the context of
sensible cities, Karima Velaquez et al. [18] describes the quality could be a key demand that ought to be
reconnoitered, permitting expedients and services to seizure info concerning the surroundings and act in the period.
During this urban quality state of affairs, within the sensible town paradigm, The Fog Orchestrator surfaces
foremost challenges, for case in point, the quality of cloud-based traffic sensing and resource management
associated with travel designing services / applications. Part of the modeling orchestration is cloud readiness and
the difficulty of managing surrounding tasks. Cloudlets deployed, bringing with them many new interesting
programming challenges. Application classification and user quality are two key aspects of programming in
providing fog and managing financial resources for their customers.

4. Edge Computing

The Edge Computing (EC) interface is the fresh from several functions. Thanks to its excellent performance
capabilities Low running costs, easy operation, high speed and excessive quality of service (QoS) in having to
deliver real-time data analysis. IoT has evolved significantly over the past two decades with pioneering advances
in evidence and telecommunication equipment.

Mukesh Prasad and others [20] explained the increasing demand from consumers, except for the high internet
value purely functional to trillions of gigabytes by edge systems. This cloud makes maximum use of dormancy
problems and bandwidth. Because out-of-date cloud servers are unable to button this massive aggregate of data
with their centralized network architecture, IoT applications require more streamlined computing management
systems in real time. Therefore, EC is essential because it overcomes the barriers of centralized infrastructure,
taking the computing capability to the edge of the link. Although the EC grasped as a good expertise, the work of
the EC is stagnant at an early stage.

4.1 Edge Computing Introduction

Edge computing since the 1990s, Akamai has released its Content Delivery Network (CDN), which executes
nodes purely nearer to end-user locations. In 1997, computer scientist Brian Noble proven how to expenditure edge
figuring through mobile devices to understand voice. This routine also used to improve cell phone series life after
two years. At the interval, this method called "imitation forges," which essentially runs the speech recognition
services provided by Apple's Siri and Google.

4.2 What is Edge Computing?

According to Edge Computing, the aging company IDC is "a network of palmtop data centers that practice or
store important data in the neighborhood and all the data collected in a footprint of less than 100 square feet.
Transferred to a crucial data center." This commonly brings up to as IoT use gears where edge expedients collect
data - often in large quantities - and send it to a cloud for dispensation. Edge attempts to compute data close by so
that some of it can handle in the vicinity, sinking traffic from the backward to the principal depository [22].

4.3 Difference between Other Methods

The data processing in a computer, on the edge computing, or on the gateway computer near the sensor.
Calculation and storage systems are located below the data generation unit, program, or component.

1324
 A Role of an Edge Computing Technologies for the Internet of Things in Smart Cities

4.3.1 FOG Computing

In fog, computing, authority-computing events are transferred to processors connected to local area network or
LAN hardware. Data in fog computing is processed in an IoT gateway or LAN within a fog node [23]. Fog
computing, first introduced by Cisco, and has emerged as a new concept that can solve latency-sensitive computing
problems. Defined as a postponement of the cloud. Fog functionality specified in Fog Servers (FS) and Foglet,
where FENs and FS are hardware nodes and Foglet Middleware for data exchange. When Fog used as a platform
for IoT, a FEN can be a neighboring smart object for complex contact and edge computing, so that knowing,
controlling besides intake can quickly achieved in FEN. In contrast to FEN dealing with interactions between smart
objects, FS focuses on the interaction stuck between FEN and the cloud. In this way, the FS reins, maintains, and
matches the FEN at its one-hop contiguity.
At the same time, Fog compromises annoyed-stage capabilities of resource observing, connectivity
management. Overall, Fog provides a cloud-to-things continuum for networking services, and facilitates
collaboration between IoT components.

4.3.2 Cloud Computing

Cloud computing acts like a standard server, responsible for all the activities in the system. Data management,
required task response, data processing, etc. are all in the cloud storage. NIST (National Institute of Standards and
Technology) provides cloud as per schedule. Cloud computing is an exemplary that allows everywhere, easy,
Service on access to a common pool of universal computing resources which can be quickly distributed and
published with marginal maintenance strength or interaction. In service providers, IoT cloud provided with the
Cloud Platform, Platform provided by the service provider as a hosting service that facilitates the development of
software submissions without the costs and hassle of obtaining and organization the underlying components
provided by the hardware and software. A novel concept of service provider hosting offered as a hosting service,
which enables the underlying hardware and software layer to disrupt the model intended to improve the information
community, promote and optimize software applications' operation without cost and hassle. Refined services by
building, exiting and developing built-in (physical and virtual) objects, providing end-to-end statistics and
communiqué technologies with universal, fast, on-demand network accessible configuration properties (e.g.,
network, server). Extensive storage, software and amenities with marginal maintenance power or profitable
outworker interactions can be easily distributed and issued Edge computing can be associated with the same data
processing (or part of the data) at the endpoint (usually a computer such as an ATM system, gateway device,
mobile, etc.) or more commonly known as an 'edge'. Initially, this 'Edge Analysis' collects streaming data, which
used to shut down the wrong ATM system or perform a task or core focuses on dividing data processing work
between edge computing and cloud-computing services to simplify IT workload and load sharing, improve the
operational capabilities of the system, and create an efficient ecosystem .for business [24]. In this table 1 the
comparison of Edge, Fog and Cloud and with future with IoT is listed. Calculation operation such as currency theft
or fatigue on its delivery device. Smart phones use this computing method when making payments through the
device; Edge Computing handles traffic lights, manufacturing like lot of activities performed by edge computing.

4.4 Edge Role in Smart City

This paper describes the Edge Computing Structure [25] for situational responsiveness in an IoT-based smart
city. IoT is an edge-computing framework for processing locations in a smart city, which helps decision makers
understand the situation and provide relevant services to city dwellers. Bin Cheng et al. [26] describes a regular-
dedicated slant to designing and implementing a hazardous fog computing-dedicated structure for fog-flow, i.e.
fog-flow, for IoT considered city stages. Limitations of the Fog Flow Programming Model IoT Real Estate
Developers can conveniently program the cloud and edges to facilitate elastic IoT stay.

In addition, it re-uses the standard interfaces of climate and reference data for housing to identify atmospheric
liveliness use in real-time - anomaly detection. The edge knob used to observer the process on or after all
operations. Once a strange case perceived on the edge, the ordinance relating to the node is resolved, and the
incident detected to process information in the cloud recorded. The system operator handles the information
through the dashboard and the structure operator can enthusiastically update the detection statute.

1325
 R.Deepa, M.Krishnamoorthi

Table 1 Comparison of Edge, Fog, and Cloud Computing and Future IoT

4.5 Edge Computing in Smart City

Zhao, N., Leung et al. [27] proposes incipient perspicacious city architecture that integrates software-defined
interacting, accumulating, and computing to ameliorate the routine of the perspicacious city applications. This
paper additionally solves resource allocation, optimization, and quandary utilizing novel, immensely colossal data
and deep reinforcement learning approach. Giordano, Andrea et al. [28] describes Rainbow middleware
architecture, which discusses three-layer architecture to providing astuteness in fog computing with Multi-agent
systems along with virtual objects to connect the authentic world sensors or physical methods.

Edge computing offers better worth of accommodation in terms of delays, power ingesting and concentrated
data traffic in the cyber world. The core feature of Edge Computing is strengthening applications that involve low
latency, location knowledge, and mobility. Ng Zhang, Hong, Zhu Zhang and others, [29] The proposed framework
eliminates difficult tracking conditions due to such closure, turning, and distortion and intensely improves tracing
accuracy. The proposed algorithm (region proposal correlation filter predicated on BACF) in constrained
recollection profile and provides more precise and robust tracking even in arduous tracking conditions such as
deformation, cluttering. Badidi Elarbi, et al. [30] describes a conceptual data integrative framework with the
heterogeneity of systems at different caliber, including data models, data semantics, accommodation
implementation, and interfaces. The architectural design to deal with issues above with edge computing, semantic
integration, and data analytics.

1326
 A Role of an Edge Computing Technologies for the Internet of Things in Smart Cities

Figure 4: Rainbow Architecture

Figure 5: Architecture for Data Integration

4.6 Challenges Addressed with Edge Computing

4.6.1. Network Bandwidth

When more information shared at the edge and further evaluation begins at the edge, the network's bandwidth
is modified. Companies traditionally allocate high bandwidth in data centers and reduce endpoint bandwidth. Now,
the edge computing challenge is the need to balance more network-wide bandwidth.

4.6.2 Distribution Information

Companies need to accept location as an additional dimension of computing. Integrated models of computing
being destroyed; Computing should now include networking as one of the main features of East-West
transportation.

4.6.3 Latency
By applying the edge to the calculation, where the proximity to the measured data is, the delay in the application
decreases with the decision delay. Moving from edge to center less and moving forward means faster responses
and faster interference and with center and edge-based calculations, application data crosses the network in every
direction, exchanging data and facing access rights. This means that data transfer is no longer a straightforward
method.

4.6.4 Security
Computing resources and software clustered in the data center so companies can standardize both digital
security and physical security. The wall can built around the capital to facilitate protection. However, Edge
Computing requires companies to implement the same network security model with more remote servers and
physical security parameters. The safety footprint issue and traffic policies are all over the city.

1327
 R.Deepa, M.Krishnamoorthi

4.6.5 Backup
Edge computing usually needed because large amounts of data can collect from different areas. Organizations
need an overlay scheme for data security that can cover this data. Because network backup does not make sense,
properties for network capacity are required for storage capacity when deciding how to store such assets.

4.6.6. Data Accumulation

Data is an important business asset, and storing it sideways brings new problems. This could create liability if
it does not comply with current data management law. Data collection and access are required, and data must be
included as part of the life cycle. As more information shared at the edge, and more assessment starts at the edge,
the bandwidth of the network will modify. Companies traditionally allocate higher bandwidth to data centers, and
reduced endpoints bandwidth.
Now, an edge computing challenge is the need to balance more network-wide bandwidth. In this section,
illustrate that the Edge computing Challenges in the table and denoted the possible solution for these identified
problems and each methods for each challenge in respectively.

Table 2 Challenges of an Edge Computing based IoT

1328
 A Role of an Edge Computing Technologies for the Internet of Things in Smart Cities

5. Conclusion and Future Work

In this paper, we have presented insight on the Internet of Things and smart city scope, various Architectures
from the initial period until the current period, details about various computing architectures, and how Edge
computing is going to plan a more significant role in smart city deployment. In addition, we have listed the
challenges in smart city deployment, and a couple of smart city architectures analyzed and discussed their key
problem and solution. However, most of the current research kinds of literature are respect to one domain or
specific problem and does not address all the critical challenges together. Finally, based on this analysis, we plan
to begin our future work.
a. Introducing Smart Middleware for Edge Computing used for any application domain with the
Interoperable Data Conversion Manager.
b. Can develop integrated collaborative middleware architecture for designing smart middleware for edge
computing architecture.
c. Plan to implement the design specifically for the use of edge, fog and cloud computing for the
development and operation of smart city applications.
d. Configuring Geospatial Maps for Sensors with Conversion Schema using Public Shareable Parameters
and Private Parameters.
e. To verify the efficiency of this design, we plan to develop an algorithm that supports Smart City
applications using an Edge Computing.
In the future, IoT is the smart city to start a novel period in smart city development. It aims to broaden the
competences of IoT and generate a common model in the design of applications for smart city architecture for edge
computing, as it intentions to completely transform the budget, civilization and our way of life.

References

A. Sethi, Pallavi, and Smruti R. Sarangi. "Internet of things: architectures, protocols, and applications."
Journal of Electrical and Computer Engineering, (2017).
B. Tiwary, Aditya, et al. "Internet of Things (IoT): Research, architectures and applications." International
Journal on Future Revolution in Computer Science & Communication Engineering 4.3 (2018): 23-27.
C. Sharma, Vandana, and Ravi Tiwari. "A review paper on “IOT” & It's Smart Applications” International
Journal of Science, Engineering and Technology Research (IJSETR) 5.2 (2016): 472-476.
D. Sarhan, QusayIdrees. "Internet of things: a survey of challenges and issues." International Journal of
Internet of Things and Cyber-Assurance 1.1 (2018): 40-75.
E. Gianni Pasolini, Chiara Buratti, Luca Feltrin “Smart City Pilot Projects using LoRa and IEEE 802.15.4
Technologies”, Sensors, MDPI, 2018.
F. Neda Sharifikheyrabadi, Mehdi Sharifikheirabadi, “Management dashboards in smart cities with a view
to the use of the Internet of Things and cloud computing”, Religacion, Center de investigations, 2019.
G. Jiong Jin, “An Information Framework of Creating a Smart City through Internet of Things” Internet of
Things Journal, IEEE, 2014.
H. Roozbehjalalli, Khalil EI-khatib “Smart city architecture for community level services through the
internet of things”, IEEE International Conference on Intelligence in Next Generation Networks, 2015.
I. Aseel Alkhelaiwi, Dan Grigoras “Data Reduction as a Service in Smart City Architecture”, IEEE
International Conference on Big Data Computing Service and Applications, 2017.
J. Federico Montori, Luca Bedogni “A Collaborative Internet of Thing Architecture for Smart Cities and
Environmental Monitoring”, IEEE Internet of Things Journal, April 2018.
K. Paul Ta-Sharma, Adnan Akbar, “An Ingestion and Analysis Architecture for IoT Applied to Smart City
Use Cases”, IEEE Internet of Things Journal, April 2018.
L. Jianhua He, Jian Wei, Kai Chen, “Multitier Fog Computing With Large-Scale IoT Data Analytics for
Smart Cities”, IEEE Internet of Things Journal, Vol.5, No.2, April 2018.
M. Bo Tang, Zhen Chen, Gerald Hefferman, Shuyi Pei, “ Incorporating Intelligence in Fog Computing for
Big Data Analysis in Smart Cities”, IEEE Transactions on Industrial Informatics, Vol.13, No.5, October
2017.
N. Pradip Kumar Sharma, Mu-yen Chen “A Software Defined Fog Node Based Distributed Block chain
Cloud Architecture for IoT”, intelligent systems for the Internet of Things, IEEE Access, 2018.
O. Pradip Kumar Sharma, Jong Hyuk Park “Block chain based hybrid network architecture for the Smart
City”, 2018 Elsevier.
P. Zaheer Khan, Ashiq Anjum “Towards cloud based big data analytics for smart future cities”, Journal of
cloud computing, Springer 2015.
Q. Zaheer Khan, Saad Liaquat Kiani, “ A framework for cloud-based context-aware information services for
citizens in smart cities”, Journal of cloud computing, Springer, 2014

1329
 R.Deepa, M.Krishnamoorthi

R. Karima Velaquez, David Perez Abreu “Fog orchestration for the Internet of Everything: State of the art
and research challenges”, Journal of Internet Services and Applications, Springer 2018.
S. https://opensource.com/article/17/9/what-edge-computing
T. Hesham el-sayed, SharmiSangar, Mukesh Prasad “Edge of Things: the big picture on the integration of
edge, IoT and the cloud in a Distributed Computing Environment”, Journal of IEEE Access, 2017.
U. https://opensource.com/article/17/9/what-edge-computing.
V. https://www.networkworld.com/article/3224893/what-is-edge-computing-and-how-it-s-changing-the-
network.html.
W. https://wire19.com/comparing-fog-computing-with-edge-computing.
X. https://www.coreitx.com/blogs/edge-computing-vs-cloud-computing
Y. SK Alamgir Hossain, Md. Anisur Rahman, M. Anwar,” Edge Computing Framework for Enabling
Situation Awareness in IoT based Smart City “, Journal of Parallel and Distributed Computing, JPDC
2018.
Z. Bin Cheng, Gurkan Solmaz, Flavio Cirillo, Erno Kovacs, Kazuyuki Terasawa, and Atsushi Kitazawa,
“Fog Flow: Easy Programming of IoT Services Over Cloud and Edges for Smart Cities”, IEEE Internet
of Things Journal, Volume: 5, Issue: 2 , April 2018.
AA. He, Y., Yu, F. R., Zhao, N., Leung, V. C., & Yin, H. (2017). Software-defined networks with mobile edge
computing and caching for smart cities: A big data deep reinforcement learning approach. IEEE
Communications Magazine, 55(12), 31-37.
BB. Giordano, Andrea, Giandomenico Spezzano, and Andrea Vinci. "Smart agents and fog computing for
smart city applications." In International Conference on Smart Cities, pp. 137-146. Springer, Cham, 2016.
CC. Zhang, Hong, Zeyu Zhang, Lei Zhang, Yifan Yang, Qiaochu Kang, and Daniel Sun. "Object Tracking for
a Smart City Using IoT and Edge Computing." Sensors 19, no. 9 (2019).
DD. Badidi, Elarbi, and Muthucumaru Maheswaran. "Towards a Platform for Urban Data Management,
Integration and Processing." In IoT BDS, pp. 299-306. 2018.

1330