Journal of Machine Learning Research 1(1) 2019 03/08/2019

Fog Computing in Cloud Computing with Internet of Things

Qasim Ali aqasim026@gmail.com
Lahore Leads University, Lahore Campus

I. Abstract

Rapid technological advancement in Internet of things and smart devices and wearable gadgets
are increasing the use of real time applications (Health monitoring, Autonomous Traffic System,
Tele-Surveillance and Live Video Streaming) which are deployed on phenomena of cloud
computing. These applications require a low latency response in order to ensure the user’s
Quality of service (QoS). To provide the low latency for above mentioned applications a new
concept is widely being used along with cloud computing which is Fog Computing. Fog
Computing has some other benefits like Low latency and location awareness, Wide-spread
geographical distribution, Very large number of nodes, Strong presence of streaming and real
time applications and Heterogeneity. In this paper we will justify that fog computing really make
possible for us to efficiently use the above mentioned applications.
II. Introduction
Cloud computing is a technique which allows computing resources as paid service on internet
and users can use these services to compute their tasks. Although it is widely being used by users
but it has some limitations [1], the most important one is distance between end user and cloud
data centers, which imposes a delay issue specifically for real time applications like live video
streaming and latency critical applications like disaster monitoring system.

Wireless Data is increasing day by day and trend toward wearable gadgets which are controlled
through cloud computing are also increasing. According to some reports by International Data
Corporation (IDC) estimated that the number of wearable devices will reach 237.1 million by
2020[2]. Furthermore, the amount of data rates generated by IoT devices will increase
drastically, as it is predicted that the total data created by a single device will reach 847 Zeta
Byte (ZB) per year in 2021, while it was 218 ZB in 2016[3]. One more reason is increase in
crowd sourced live video streaming at maximum rates of multi-terabits per seconds. According
to Cisco mobile video traffic predictions, in 2021 video content will be 82% of the global
Internet traffic. In 2021 the network will be crossed by around 1 million minutes of video content
every second. This kind of increase in data and devices will become major cause of issues to
cloud computing as it needs to ensure low latency and support high mobility. Fog computing is
the promising solution that augments cloud computing by providing computational and
networking resources to the proximity of end users.

III. Fog Computing

Fog computing was announced by the Cisco in 2012, and it is defined as” Fog Computing is a
highly virtualized platform that provides compute, storage, and networking services between end
Journal of Machine Learning Research 1(1) 2019 03/08/2019

devices and traditional Cloud Computing Data Centers[4], typically, but not exclusively located
at the edge of network.

IV. Fog Computing VS Cloud Computing

Parameters Cloud computing Fog computing

Server nodes Location Within the Internet At the edge of the network

Client and Server Distance Multiple hops Single/Multiple hops

Latency High Low
Delay Jitter High Very low
Security Less secure, Undefined More secure, can be defined
Awareness about Location No Yes
Vulnerability High Probability Very low probability
Geographical Distribution Centralized Dense and Distributed
Number of Server nodes Few Very Large
Real Time Interactions Supported Supported
Kind of last mile connectivity Leased line Wireless
Mobility Limited support Supported
Table 1: Comparison of Cloud Computing and Fog Computing

Table shown above provides us the comparison among the cloud computing and fog computing
by using list of parameters like latency, security and geographical distribution etc.

V. Architecture
Journal of Machine Learning Research 1(1) 2019 03/08/2019

Fig 1: Architecture of fog computing in cloud computing environment

As from the above given diagram you can see cloud computing system is divided into different
layers like cloud layer, fog layer and edge layer. In cloud layer there are different data centers
and servers and in fog layer which is closer to end devices for fast computing purpose, fog layer
contains multiple gateways by which multiple end devices(IOT devices) are actually connected
to the clouds.

Fig 2: Fog Layer

As cleared from the above diagram which gives us the clear picture of fog layer all the gateways
and switching nodes are located on the fog layer and all the IOT devices actually make their
communication with data centers possible with help of fog nodes like gateways.

VI. Challenges for Fog Computing

a) Privacy
Journal of Machine Learning Research 1(1) 2019 03/08/2019

As the fog computing totally rely on the wireless communication so the major concern is
privacy. As the fog nodes are deployed at the edge of the internet so the massive maintenance is
required for edge nodes[5].

b) Security
As the fog layer contains multiple gateways so the authentication of nodes is a major concern.
Each node has its own IP address. A malicious user may use a fake IP address to access
information stored on the particular fog node. To overcome this access control an intrusion
detection system has to be applied at all layers of the platform[6].

c) Placement of Fog Servers

Major issue is to decide the location of fog servers in cloud computing to fulfill the requiems of
end users. Analyzing the work done in each node in the server before placing them reduces the
maintenance cost.

d) Delay in Computing
Delays due to Data aggregation, Resource over-usage reduces the effectives of services provided
by the fog servers, causing delay in computing data. Data Aggregation should take place before
data processing, Resource-limited fog nodes should be designed scheduling by using priority and
mobility model.

e) Energy Consumption
Since fog environments use large number of fog nodes, the computation is distributed and can be
less energy-efficient. Hence, reduction of energy consumption in fog computing is essential [7].

f) Management Effort
As in fog computing all the data and computation is closer to end user, so in dense fog
deployment there will be very large number of fog nodes, and they need massive management
when scaling, installing updates or updating the configurations. There is not already managed
fog infrastructure service so this management effort is left with application providers.

g) Managing Qos
As the no of nodes increases complexity of systems also increases because of nodes geographical
distribution. This is due to issues and faults (ranging from simple ones such as network latencies,
message reordering, or message loss to complicated faults such as network partitioning or
byzantine failures). Usually, systems can cope with these issues but they still affect QoS and
their tradeoffs.

h) Physical Security
Fog nodes are deployed near to end user so there are chances of intrusion of data by the intruders
when data is moving through fog layer.
i) Fog Service Providers
Journal of Machine Learning Research 1(1) 2019 03/08/2019

While our research is unlikely to directly result in fog infrastructure services, there are still a
number of open research challenges that potential fog service providers need to face once they
decide to enter the market.

j) Capacity Management
One of the main properties of cloud services is the illusion of infinite resources [8]; for edge
nodes, this illusion becomes impossible. Essentially, in an edge node there is a limited set of
resources that is available to fog service consumers. In the cloud, resource pooling across a large
customer base and inside gigantic data centers helps cloud providers even out natural demand
variability. In Edge Computing, the limited capacity of nodes does not allow this: edge nodes
will alternate between states of idleness and periods with more demand than they can handle.
However, offloading processing and data to adjacent nodes or intermediary nodes is bound to
impact QoS. This leads to the following three main questions for fog service providers:

VII. Applications of Fog Computing

a) Mobile Big Data Analytics
In IOT (Internet of things) data is collected in bulk and it can’t be stored in cloud (Not
efficient
Data cannot be stored in cloud as the IOT data is collected in bulk. In such situations it is
beneficial to use fog computing instead of cloud computing as fog nodes are much closer to end
systems. It also eliminates other problems such as delay, traffic, processing speed, delivery
time, response time, storage data transportation and data processing. Fog computing could be
the future of IOT applications.

b) Water Pressure at Dams

In Dam System Sensors installed in dams send data to the cloud where the data is analyzed and
officials are alerted in case of any discrepancies. The problem faced here is the delay of
information which could be dangerous. To solve this, Fog is used, and since it is near the end
systems it is easier to transmit data, analyze it and give instantaneous feedback.

c) Smart Cities
Concept of smart city is much attractive for every one as make the life of everyone smooth and
easy. A Smart City is an urban area where various sectors collaborate to achieve sustainable
results by analyzing real-time data [9]. Building smart cities imposes the challenge of ensuring
accurate quick response when monitoring the health of the infrastructure building blocks such as
oil and gas pipelines, subways and roads. In addition, the massive data generated by the sensors
leads to big data analysis issue. The authors in [10] introduced a fog computing architecture for
big data analysis in smart cities, they analyzed a case study of a smart real-time pipelines
monitoring system, that detects any hazardous events such as leakage, corrosion or failure using
machine learning algorithms. The proposed architecture is composed of four layers. First, the
optical fibers that senses the temperature of the pipelines. Second, the fog nodes that detect any
threats using machine learning algorithms, and performs features extraction to send the results to
Journal of Machine Learning Research 1(1) 2019 03/08/2019

the upper layer for further analysis. Third, the intermediary computing nodes incorporate spatial
temporal association for event recognition. One of the smart city is built in Qatar and Saudi
Arabia is also going to built that kind of smart city.

d) Smart Transportation
Smart transportation is a concept of autonomous transport in which there is no need of driver and
vehicle automatically decides the direction in which it needs to move. One of the autonomous
transport system is launched in Dubai and china is also going to launch that kind of
transportation system. Fog computing is a major technique which is used in those kind of
transportation to send the data to data centers and compute to analyze.

e) Tele-surveillance
In gunshot areas and in shopping malls where theft need to be avoided can be achieved by
placing the CCTV cameras near to fog nodes a special kind of video content software is used to
identify the critical conditions.

f) Health Care
Now a day in health care system cloud computing concept is used for both diagnostic system and
maintaining the data of medicines, patients and staff of hospitals. Quick and accurate reporting of
activities in health care system is made possible by using the concept of fog nodes for analyzing
the disease reasoning and remedy for disease. In health care department of Punjab cloud
computing with fog computing is used for all above mentioned reasons.

VIII. Benefits of Fog Computing

a) Low latency
One of the best benefit of fog computing is it provides the low latency in cloud computing
because fog layer is closer to end users.

b) No problems with bandwidth

Pieces of information are aggregated at different points instead of sending them together to one
center via one channel.

c) Loss of connection is impossible

In fog computing there are multiple channels so chances of connection loss is impossible.

d) High security

e) Improved user experience

There is no downtime in fog computing that’s why user is much more satisfied.

f) Power-efficiency
Edge nodes run power-efficient protocols such as Bluetooth, Zigbee or Z-Wave
Journal of Machine Learning Research 1(1) 2019 03/08/2019

IX. Conclusion
Purpose of this research paper is to reduce the problem of cloud computing which high latency
time by introducing a new concept of fog is computing. Here short comparison of cloud
computing and fog computing is also provided. Moreover, this paper highlighted the potentials
of fog computing, which are mobility support, minimizing the latency, reducing the load on the
cloud, data centers computation offloading and energy efficiency. In addition, a detailed fog
computing architecture is explored. Twelve recently proposed works for applications in smart
cities, smart transportation, tele-surveillance, health care and multimedia are reviewed. Finally
the benefits of fog computing is also provided, so that people can do more research on fog
computing.
X. References
I. [1] M. Firdhous, O. Ghazali, and S. Hassan, “Fog computing: Will it be the future of
cloud computing?,” p. , The Third International Conference on Informatics &
Applications (ICIA2014), 2014.
II. [2] “Idc forecasts shipments of wearable devices to nearly double by 2021 as smart
watches and new product categories gain traction.” Accessed on 26 May 2018.
III. [3] “Cisco global cloud index: Forecast and methodology, 2016-2021.” Accessed on 18
May 2018.
IV. [4] F. Bonomi, R. Milito, J. Zhu, and S. Addepalli, “Fog computing and its role in the
internet of things,” in Proceedings of the first edition of the MCC workshop on Mobile
cloud computing, pp. 13–16, ACM, 2012.
V. [5] Shanhe Yi, Zhengrui Qin, and Qun Li, "Security and Privacy Issues Of Fog
Computing: A Survey", College of William and Mary.
VI. [6] Shanhe Yi, Zijiang Hao, Zhengrui Qin, and Qun Li, "Fog Computing:
Platform and Applications", dept. Of Computer Science, College of William and Mary,
IEEE - 2015.
VII. [7] Amir Vahid Dastjerdi, Harshit Gupta, Rodrigo N. Calheiros, Soumya K.
Ghosh, and Rajkumar Buyya, "Fog Computing: Principles, Architectures, and
Applications", IEEE -21016.
VIII. [8] Shanhe Yi, Zijiang Hao, Zhengrui Qin, and Qun Li, "Fog Computing:
Platform and Applications", dept. Of Computer Science, College of William and Mary,
IEEE - 2015.
IX. [9] C. Perera, Y. Qin, J. C. Estrella, S. Reiff-Marganiec, and A. V. Vasilakos, “Fog
computing for sustainable smart cities: A survey,” ACM Computing Surveys (CSUR),
vol. 50, no. 3, p. 32, 2017.
X. [10] B. Tang, Z. Chen, G. Hefferman, S. Pei, T. Wei, H. He, and Q. Yang,“Incorporating
intelligence in fog computing for big data analysis in smart cities,” IEEE Transactions on
Industrial Informatics, vol. 13, no. 5,pp. 2140–2150, 2017.