See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/351774895

Green Computing in Cloud Technology

Article in Linguistica Antverpiensia · March 2021

CITATION READS

1 2,833

1 author:

Pravin Kshirsgar
Avn Insitute of Engineering and Technology
33 PUBLICATIONS 31 CITATIONS

SEE PROFILE

All content following this page was uploaded by Pravin Kshirsgar on 22 May 2021.

The user has requested enhancement of the downloaded file.

 LINGUISTICA ANTVERPIENSIA, 2021 Issue-1
www.hivt.be
ISSN: 0304-2294

Green Computing in Cloud Technology

Pasupuleti Sailaja, Research Scholar, KLEF; Email: [email protected] Dr. P.
Lalitha Surya Kumari, Professor, KLEF

Issue Details Abstract

Issue Title: Issue 1
Received: 15 January, 2021 According to Greenpeace, one of the major contributors of climate conditions change is
Accepted: 08 February, 2021
Published: 31 March, 2021 the consumption of the fuel and gas, which results in atmospheric pollution such as
Pages: 2797 - 2806 carbon dioxide (CO2). Cloud storage is made up of data centres spread all over the
world, and it uses a lot of resources to operate. Green Cloud Computing is a model
Copyright © 2020 by author(s) that has the benefits of being environmentally sustainable, energy consuming,
andLinguistica Antverpiensia
consolidating, virtualizing, and so on. When the use of cloud computing expands, so
does its effect on the world. Collaboration between software and hardware software
partners, as well as cloud providers, is necessary for green computing in cloud
environments. The aim for this particular paper is given that to examine the past green
computing and the cloud computing. It also providesabout cloud technology and
current green computing systems that were applied to cloud platforms. In addition,
related work is discussed, which offers an overview of new cloud technologies as well
as green computing strategies. This article also discusses the challenges,
considerations and requirements that cloud data centres and cloud organisations must
understand as well as certain aspects that affect green cloud use. Finally, the proposed
conceptual architecture for green computing on cloud is presented

Keywords
Green computing, Cloud Computing, Server, Virtual Machine, Physical Machine,
Service, Model, Energy, Framework.

I. INTRODUCTION
Cloud computing has emerged as the go-to solution for on-demand applications, service
aggregation, and computing technology around the world. The powerful implementations,
enhancements in the cloud development and useful great services providing by many cloud
computing companies, for example AWS, Oracle, Amazon, Microsoft Azure, etc. These
factors also aided cloud computing's success. Many customers may be able to use cloud
storage (Figure 1) to meet their complex requirements while their in-house IT technology is
unable to meet them. Any cloud data infrastructure is a remote access method to any data
centre that is situated anywhere other than the customer's physical premises and allows the
customer to access their services, computers, or data over the internet. The cloud service
does routine repairs and updates to secure the data during outages and other faults, and it
also owns numerous data centres in multiple geographic regions. Using sophisticated
“Artificial Intelligence” and “Machine Learning” algorithms, for example, to recommend
products that based on complicated user specifications. Furthermore, many start-ups and
companies are switching from an on-premises datacentre to the cloud storage in order to
minimise or remove skill needs and IT running costs. For example, Expedia, Vodafone,
Netflix, etc. are the large enterprise that use services of cloud computing. If the world shifts
to the cloud storage, cloud providers must proactively and continuously broaden their
datacenters in order to provide their best capabilities while still enhancing maintenance and
management without sacrificing service efficiency.

Pasupuleti Sailaja is Research Scholar in Koneru Lakshmaiah Education Foundation, Hyderabad-500075, Telangana State, India (e-mail:
[email protected]).
Dr. P. Lalitha Surya Kumari, Professor in Koneru Lakshmaiah Education Foundation, Hyderabad-500075, Telangana State, India (e-mail:
[email protected]).

2797 LINGUISTICA ANTVERPIENSIA

 LINGUISTICA ANTVERPIENSIA, 2021 Issue-1
www.hivt.be
ISSN: 0304-2294

Figure 1: Cloud Computing

When the use of cloud computing grows, so does its effect on the climate as the need for
electricity grows. Everywhere, including cloud computing, increased energy demand has
become a challenge. As per the article "The Road to Energy Saving in Cloud Computing"
written by Tony Mastelic, “the current power consumption of cloud computing has exceeded
the value 1% of the total global electricity consumption”. Hence, the importance should be
given to sustainability [10]. Furthermore, it explores that we can fulfil our requirements
while decreasing our negative effects on climate and the upcominggeneration lives. It came
out an important and top-of-mind issue in today's environmental world. Cloud vendors must
bear certain burdens in terms of ensuring environmental protection. Consequently, the topic
of cloud computing in the world has been moderated by green computing [11]. Green cloud
infrastructure focuses on environmental issues that are relevant and impactful. According to
the results, in recent years researchers have paid more attention to green cloud computing.
Green computing is also environmentally sustainable and can again defined due to the usage
of the computing resources. Furthermore, it eliminates environmental risks by integrating or
retaining resources for customers, such as cloud vendors, businesses, and end-users. A great
deal of cooperation between service vendors and their applications and hardware suppliers is
anticipated to accomplish green computing in cloud environments. For example, hardware
partners should produce environmentally sustainable power-consumption systems, and
computerised applications to minimise power consumption should be created by software
partners. Cloud vendors can also take part in different strategies such as energy-efficient
deployment, pollution reduction, server virtualization consolidation, and recycling, to name a
few.
There are many algorithms for many requirements and performing in cloud environment.
Generally, the search process is complex in the cloud environment for further maintenance
and required a good optimization performance. Task planning is an NP-complete problem in
cloud environments. And intelligent heuristic algorithms are commonly considered to be
superior solutions to NP problems. Mostly the Heuristic (Figure 2) intelligent algorithms
have been used for resolving tasks by scheduling requirements in the cloud computing.
Currently, there are many resources scheduling strategies and these are based on “Ant
Colony Optimization” (ACO), “Genetic Algorithm” (GA) and “Particle Swarm
Optimization” (PSO).

Figure 2: Heuristic Search Algorithm

The adoption of cloud computing in Saudi Arabia was fraught with difficulties. Privacy
considerations, stability, adherence to government policy, loss of power, and a lack of
expertise are all part of it. Privacy and stability are also two key factors shaping Saudi
Arabia's adoptions and usage of the cloud computing. Increased protection and privacy risks
as information and data are stored on Internet Web pages. Data stored in online systems are

2798 LINGUISTICA ANTVERPIENSIA

 LINGUISTICA ANTVERPIENSIA, 2021 Issue-1
www.hivt.be
ISSN: 0304-2294

possible to access with malicious intent by the people, possibly leading to secrecy and
recording being manipulated. Secondly, the Saudi Arabia government has stated that its
policies on the usage of the public/community cloud computing are inadequately
implemented, making it impossible to incorporate cloud computing. Ebtesam AlharbiH,
Sahar S,Maryam M. Alahrbi, Alkhamali, [1] have taken into consideration the version of
inexperienced computing in a Saudi Arabia. And, it proposed the framework that illustrated
the elements that have an effect on the version of inexperienced computing for cutting-edge
and destiny of the cloud computing carriers in Saudi Arabia. Also, the system should be
seen as guiding principle in ensuring that novice programming is completed in a professional
manner. Their paper is divided into several parts, one of which requires them to illustrate
works and responses directed at beginner computer users. In some other section, elements
influenced on adoptions inexperienced cloud computing, associated problems for
information facilities and cloud corporations had been taken into consideration and
discussed.

In the UAE, S. Paul [2] conducted a small survey to gauge employee knowledge of green
cloud computing. The questions concentrate on the environmental effects of green
computing and the readiness of workers to enforce green in computing principles and
strategies. It also displays some of green cloud modelling processes to demonstrate how
reliability of green computing is accomplished.

The work done by S. Garg, R. Buyya, [3] and They've proposed a green cloud system to
reduce cloud data centre carbon emissions and energy usage. They use the green broker to
submit demands and then identified the suitable cloud datacentres [19] to meet them. For
each user request, they have built a framework to monitor energy usage. In addition, the
greenhouse emissions measurement and the expense of each cloud service will be used. The
green broker is required to choose cloud providers and track them based on the quality-of-
service criteria of the users. It also ensures that carbon emissions are kept to a minimum to
benefit all users.

“Mohammad Shojafar, Claudia Canali, Riccardo Lancellotti and Jemal Abawajy” [4] have
mentioned according to their paper, cloud data centres are becoming increasingly energy-
intensive and efficient, posing a significant challenge for both economic and environmental
reasons. Improved energy quality in cloud data centres necessitates a strategic solution. And
it can use traffic engineering methods to automatically adjust the number of active servers
based on the existing workload. To that end, they've proposed MMGreen, a joint computing
platform that includes connectivity enhancement and takes advantage of virtualization
technology. They explored in their proposal the conventional multimedia processing
scenario with a wide range of data exchanges and intense computational tasks. The proposed
architecture guarantees not only quality of service (QoS) for consumers, but also optimises
green cloud storage objectives and energy conservation with the maximum distribution of
CPU frequencies and is built upon DVFS. The authors used MMGreen in both virtual and
real-world workloads to evaluate the proposed framework's actual efficacy. According to the
results of MMGreen's experiments, it significantly reduced the expense of energy for
networking, computing, and reconfiguration as opposed to SLA-compliant resource
provisioning approaches.

They suggested the MM novice optimization method in their paper for a shared flexible
allocation [13] of computational and communications fees in power-green records facilities
for multimedia statistics processing. The poorly coupled form of the resulting optimisation
problem was exploited in order to find an objective solution since it is naturally non-convex.
In both artificial and real-world workloads, our tests show that MMGreen outperforms any
other alternative for a DVFS-compatible recording centre, saving up to 60% in power over
traditional static solutions in recording centres that haven't adjusted the CPU's operating
frequency to match their input workload. Furthermore, our comprehensive performance
analysis reveals the impact of algorithm/logic and version parameters on the overall impact
of performance.

Zhong Zong, JSTI Group, Nanjing [5] have intended to use as little resources as possible
while still shortening the mission's execution time regarding the green cloud computing.
Their paper suggests a complex fusion programme algorithm that integrates an ant colony

2799 LINGUISTICA ANTVERPIENSIA

 LINGUISTICA ANTVERPIENSIA, 2021 Issue-1
www.hivt.be
ISSN: 0304-2294

with genetic algorithms. As a result, the amount of power used in storage and cloud
computing data centres is negligible. According to simulation results, the proposed algorithm
for preparing work reduced the duration and the total energy consumptions of the cloud
computing operations significantly.Their research focuses on green cloud infrastructure,
which seeks to reduce cloud computing systems' energy use while increasing electricity use.
It proposes a set of rules for project scheduling that include ACO and
GA[21][22][23][24][25][26[27]. In terms of assignment execution time and overall power
consumption, it compares GA and ACO. The results indicate that the set of rules can be used
to increase strength.

Heng Liu, Xiaofen Zhang, Jing Bi [6] They worked on a number of projects in various
GCDCs, including job scheduling, service quality agreements, mission failure probability,
service rates, electricity prices, and clean energy. To forecast solar and wind energy, they
initially relied on a long short-term memory network. The corporation then used a bipolar
modelling algorithm to improve GCDC's sales-to-costs ratio. Finally, real-world data is
included, such as wind power, trace working load and solar energy. The efficiency of the
suggested work scheduling procedures and electricity estimates were also demonstrated by
the use of electricity.It was revealed that the suggested approaches outperform other neural
network methods.

Green energy sources are being investigated and tested for the construction of green cloud
data centres. It is difficult to make an accurate renewable energy forecast, so good activity
planning practises should be established. To emphasise the importance of forecast records
for job scheduling precision, this paper proposes a solution that incorporates LSTM-based
estimations and bi-target optimization. Experiments in the real world show that the LSTM
can correctly predict solar and wind speed patterns, making it the perfect basis for intelligent
project planning in GCDCs.In addition, the findings of bi-goal optimization showed the
significance of the GCDC results to help plan various bond forms.

Abdulaziz Alarifi, Kalka Dubey, Mohammed Ammon, Torki Altameem and team [7] have
concentrated on the increased need for cloud computing services as a result of the cloud's
increased elasticity and digital transformation, which necessitates increased hard work to
advance the energy of electrical performance of the cloud data centres. In their paper, they
have tested the framework of an EnergyEfficiency Hybrid (EEH). The aim of this scheme is
to increase the productivity with which data centres use electrical resources. Rather than
focusing entirely on one approach, as in previous related works, the new framework
integrates server restructuring as well as request timing.Until scheduling, the EEH system
sorts the clients' demands (tasks) based on their time and energy requirements. It has an
algorithm for planning, which takes care of energy consumption once taking decisions of
planning. It also represent a consolidation logic/algorithm, which makes it necessary to sleep
or hibernate servers, crowd servers, migrate digital machines, and migrate servers.
Furthermore, the EEH architecture has a series of migration guidelines for moving migrated
virtual machines to new servers. In the terms of electricity use effectiveness (PUE), statistics
centre energy productiveness (DCEP), common elapsed time, cost savings and throughput,
simulation studies showed that the EEH method outperforms the use of a single
methodology for reducing energy consumption.

This research painting suggested and tested a hybrid architecture for novice cloud
computing that takes into account the time-based totally energy usage paradigm. The
proposed framework is focused on the use of both scheduling and restructuring methods in
addition to the other approaches proposed in literature. In the beginning, customer
requirements are handled according to both power and service time requirements. The most
fitting VM for each request is then assigned based on a recommended algorithm for
programming. Thereafter, the merged servers and servers that accept VMs from unified
servers are added under a pre-defined set of merger guidelines. Finally, to complete the VM
migration from the combined servers, a replication algorithm is used.The EEH approach is
based on only one method, which reduces energy consumption by PUE, DCEP, common
implementation time, throughput and cost savings. The EEH method has a better
performance. Failure results on the total amount of used energy will be investigated in our
future experiments. Deep learning models may be used to forecast server use and examine

2800 LINGUISTICA ANTVERPIENSIA

 LINGUISTICA ANTVERPIENSIA, 2021 Issue-1
www.hivt.be
ISSN: 0304-2294

different scheduling and restructuring requirements. We also plan to incorporate a load

balancing strategy into the scheduling algorithm.

Christos StergiouL., Thessaloniki, Kostas PsannisE.,Greece, Yutaka Ishibashi [8] They

investigated and surveyed the unresolved issues in the fields of green cloud infrastructures
and energy-efficient computing. CloudSim (Figure 3) is a helpful piece of software for
comparing Cloud environments. It has been used to introduce and clarify their definition. In
order to achieve their goal, they used the CloudSim simulator architecture. Then, they
researched and suggested a system architecture for improved big data processing built on a
cloud federated network. They also suggested a Big Data Optimization Algorithm in the
Green Cloud world to accomplish an energy-efficient distribution of resources[15].Their
findings showed that their proposed model has a lot of promise, as it provided substantial
efficiency benefits in terms of improved data storage and cost savings even under high
workload scenarios.

Figure 3: CloudSim Simulator

Since they found that this CloudSim simulator was the right simulator of cloud for their
needs, they examined related field works trying to develop a powerful green and the green
cloud platform for the data storage and itsprocessing with the simulator of CloudSim. The
CloudSim can opt to include components such as VM and CCo, which can be configured
and used in a case ofrealtime infrastructures.It is based on their findings, they've suggested a
system architecture based on Cloud federated culture for increased use of large records
control.They tried to incorporate in their situation a set of instructions via the proposed
single version to achieve an energy-efficient useful resource distribution approach in the
green cloud world for big data management. They want to incorporate the most pressing
issue of security into the proposed architecture in the future, with the intention of providing a
green and comfortable Cloud ecosystem through federated learning and reinforcement
learning scenarios.

Ahmadreza Montazerolghaem, Mohammad Hossein Yaghmaee [9] have shown the rapid
expansion of multimedia network networks and collaboration has been stated. They cited
Voice over Internet Protocol (VoIP) as an example of multimedia network services that have
caused resource shortages in these networks from two perspectives: 1. Burden due to a lack
of the resources; 2. Power loss as a function of resource redundancy According to them,
cloud computing platform allowing the scale of capital needed to increase or decrease in
response to customer demand. Virtualization technology and resource sharing is responsible
for many of the cloud computing standards.In contrast, the latest SDN Model gives an
overview of the entire network for automatic resource management. It's not the only way to
handle the network. In her paper, they suggested GreenVoIP, a power-efficient infrastructure
management tool for virtualized VoIP cloud centres. And the network architecture and
number of VoIP servers like switches can be managed. It not only avoids surplus loads, but
also encourages solar computing through energy savings. GreenVoIP is a novel system that
they developed, applied, and tested for their research.Based on NFV and SDN Technologies,
the resource management technique was used to balance resources between the VoIP and the
overload network. In this case, SIP-VNFs display the SIP server functionality graphically.
Instead of traditional switches, openflow switches are also used within network networks.

2801 LINGUISTICA ANTVERPIENSIA

 LINGUISTICA ANTVERPIENSIA, 2021 Issue-1
www.hivt.be
ISSN: 0304-2294

The management and maintenance of SIP-and VNF's OpenFlow switches effectively

eliminates overload and limits the number of active devices in the network.They've
mentioned approximately upcoming work that includes more precise look at the parameters
brought of their present-day article, together with α, and the look for their most appropriate
values. Additionally, they deliberate to learn the effects of the parameters on each other. The
“θ” and “ϕ” threshold value is vital inside the trade off among overloading the incidence and
strength consumptionsof their gift take a look at, the thresholds are considered to be
constant. They cited that for future work, clever Fuzzy techniques [12] might be offered to
exactly decide those thresholds. They mentioned that although the current architecture can
be used and tested for other forms of networks, it is designed for VoIP networks. In addition,
future progress on their paper will include the development of an excellent mathematical
model for predicting SIP-VNF migration. Constraints, connection delays, and network
provider-level settlement will also be part of their model (SLA). They also intended to use
gadget learning to forecast server loads.

II. PROPOSED METHODOLOGY

A. Cloud Computing

“National Institute of Standards and Technology” (NIST) definesthe cloud computing

such as a " Cloud computing is a model for enabling ubiquitous, convenient, on-demand
network access to a shared pool of configurable computing resources (e.g., networks,
servers, storage, applications, and services) that can be rapidly provisioned and released with
minimal management effort or service provider interaction.” A wide range of large-scale and
small-scale technologies can be found in the cloud. Users have nearly unlimited options for
the number and types of programmers they can deploy to the cloud. Many tech firms are
already migrating their mission-critical systems from costly internal data centres to Cloud
Solutions. Cloud-based applications are typically very flexible. A resource may be manually
added or removed by an application administrator. The programmers can be set up to
automatically scale.The cloud system consists four implementation models, three service
models with supporting of many key features. Figure 4 explains the framework of
cloudcomputing [16].The applications include self-pooling, by the demand service, fast
elasticity, massive networking and computed service. Service application software (SaaS),
device platform (PaaS) and service framework provide cloud service models (IaaS). Models
for deployment cover the public cloud, private cloud and mixed cloud.

Figure 4: Cloud Computing Framework

SaaS states the ability granted to the user to use the provider's software when working on
the cloud platforms (Software as a Service). The applications can be accessed through a
thick or thin client interface from a number of client devices. PaaS (Platform as a Service)
refers to the development and implementation of software on a cloud, which contains
application servers, databases, and other elements delivered like a service that to provide
anything they want to build their application using the provider's services, libraries,
programming languages, and the tools.IaaS (Infrastructure as a Service) is a “cloud

2802 LINGUISTICA ANTVERPIENSIA

 LINGUISTICA ANTVERPIENSIA, 2021 Issue-1
www.hivt.be
ISSN: 0304-2294

networking infrastructure concept that refers to the ability to provide clients with storage,
provisioning processing, networks, and a host of other simple computing resources”. Any
cloud data centre that follows the standards of service models and their characteristics can be
classified as a cloud computing type. The cloud data centre should be able to satisfy on-
demand computing demands without requiring human involvement, such as provisioning
more disc space, CPU power, and other resources. This should allow its services must be
accessible through the network, i.e. the public Internet. Resource pooling that is really
necessary to ensure that the data centre meets the various supply and scalability needs of
many customers in a smooth and fast way. Furthermore, the use of cloud providers can
continuously be calculated and thus accurately reckoned that consumers are paying for their
use. In addition, the service model should allow customers to choose appropriate services in
accordance with their needs. SaaS (Software as a service) provides a range of free and
paying cloud-based software in various cloud data centres providers.For example Google
Apps.SaaS iscompletely controlling by its owners that are providers. PaaS is allowing
consumers for developing and implementing their applications using PaaS platforms. For
example, database servers, language services, tools related services, etc. IaaS allows users to
fully monitor the provision and virtualization of the majority of their infrastructure[17]. For
example, storage, servers, network, etc. There are four different ways for deployment in
Cloud computing. The public/community clouds are the most common deployments of the
cloud platform, it designed for the purpose of public usage, and it is accessible via the public
network. The cloud customers/end users of public cloud are charging based on the usage of
services, resources, etc. The private cloud is a particular computer service,which provides
consumer companies with controllable, stable and customised networks. Any customer
organisation can then receive networks that are committed to its particular needs. Hybrid
clouds incorporate benefits of public and private clouds in terms of data security and the
ability to scale their IT architecture when required. The Cloud community enables a network
of organisations, such as improving the safety of banks requiring computer technology, to
collaborate together on similar projects. Cost of the cloud group, designed on public and/or
private clouds, is shared by users.

B. Green computing

According to Cisco, cloud service systems will account for about 94 percent of
computation by 2021. Furthermore, the International Data Corporation (IDC) forecasts that
by 2025, the volume of data produced and used would have increased to 175 zetta bytes.
This necessitates the installation of new data centres and services by cloud providers. As a
result of the types of facilities and appliances, more enterprise storage centres and networks
are provisioned in cloud. It is resulting in higher electrical demand. Customers can access
cloud storage services by VMs (Virtual Machines) that are built and runin data centres.The
data centres are made up of several physical servers [14], each with its own collection of
services. As a result, every cloud contains a huge number of materials that consuming a lot
of electricity, resulting in a lot of CO2 emissions. While cloud data centre energy use
appears to be the primary driver, several other factors are also at play, including waste and
recycling, the lack of green energy use, inappropriate data centre architecture, and data
centre location. Many leading cloud providers (such as Google) have adopted green
computing in their data centres as a result of these issues, including the use of solar energies
and personalised evaporative cooling.

Figure 5: Green Cloud Computing

2803 LINGUISTICA ANTVERPIENSIA

 LINGUISTICA ANTVERPIENSIA, 2021 Issue-1
www.hivt.be
ISSN: 0304-2294

Green Cloud Computing (Figure 5) is a model that has the benefits of being
environmentally friendly, energy efficient, converged, virtualized, and so on. It benefits not
only the climate, but also the cloud data centre and cloud service provider because it makes
better use of resources. The primary drivers of the need for green computing are biodiversity
and climate change. The global atmosphere is shifting fast, according to Greenpeace,
resulting in a variety of environmental problems, including increasing levels of water,
elevated storm frequency, and melting Antarctic and Arctic ice. Greenpeace is a NGO that
advocates environmental conservation and sustainability in the long term. Greenpeace said
the use of oil/fuel and gas such that they lead to atmospheric pollution such as carbon
dioxide, is one of the major causes for contributing to climate change (CO 2). Any cloud
storage system is made up of data centres, which use a lot of resources to run. Energy is
frequently derived from power providers that depend heavily on gas and oil.As a result,
green computing has emerged as a key issue that any cloud provider can solve in order to not
only protect the environment but also to increase overall efficiency and cut costs. Green
Cloud computing can be an environmentally sustainable model because it protects the world
from being harmed by cloud computing and promotes sustainability. It can also reduce cpu
load and reuse energy, allowing for more stable and reliable resource use.

C. Proposed Framework for Cloud Computing

Every cloud provider has one or more cloud data centers and they deploy these in different
Geographical region. And each data center has many Virtual Machines that can help or
serve customer requests. The customers requests are again for different purposes such as
data retrieval requests, data modification requests, service update requests, service resources
allocation or de-allocation requests, service requests for batch executions, migration
requests, etc. Some of them require less resources for processing the requests comparing
with other requests i.e., energy consumption is relative depend on service request type. For
example, service requests for batch executions require more resources for processing.
Moreover, arranging more cooling systems to maintain temperature at cloud data centers is
again cause for additional energy consumption due to more cooling systems. From all these
perspectives, each cloud data center should consider are these points and environments
concerns as well. If any cloud data center at Geographically cooling location, then cooling
systems can be avoided and these cloud data centers can be utilized for service requests of
batch processing such as backup jobs, orchestration, etc. Hence, need a frameworkwhich
should be eco-friendly for cloud data centers. The framework (Figure 6) should consider the
cloud service request purpose and Geographic region of cloud data center and finally it
should be solution for power consumption problem. At every cloud data center can be found
that the physical machines or physical servers or super cluster servers and these are
benefiting for virtual machines deployments.

Figure 6: Proposed Framework for Green Cloud Computing

The important objective of the green cloud computing is that the reduction of the amount
of the total electrical power consumption by type of service requests. Based on service
subscriptions and/or based on service purpose by customer, service type can be predicted.

2804 LINGUISTICA ANTVERPIENSIA

 LINGUISTICA ANTVERPIENSIA, 2021 Issue-1
www.hivt.be
ISSN: 0304-2294

Also, with the help of metric collections about cloud service usage, service type can be
predicted. These can be input for understanding the energy consumption for those services.
In further it benefit to decide to run the high-energy consumption service requests at
geographically fit cloud centers without more cooling systems. Also power efficiency can
be calculated at each cloud data center using two values. That are (a) “the amount of
electrical power consuming by all components located in the cloud data center, here the
components are storage resources, computing resources, cooling devices, generators and
networking devices and (b) the amount of electrical power consumed by all IT components
located in the cloud data center, here the components are storage, computing and networking
devices”. And it is mandatory to calculate these amounts periodically to predict and find
more probability values. Cloud migrations by customers and migration activities at cloud
data center as part of maintenance activities by providers can result more energy
consumption. In this consideration, the time of the migration need to be calculated and this
can be calculated using the amount of the data involving in the migration with available
bandwidth of the migration activity.

Customers' support requirements, as well as their best standards for and operation, are
added to the cloud data centre. A request for cloud service is a request that is sent by a VM
in the data centre of the information cloud. Customers with special processing power
specifications and software instances permit demands for services. The processor rate,
memory power, failure rate, current usage percentage, energy intake price, availability and
several other parameters on this device must be stored for every VM (Virtual Machine) or
any physical resource. The scheduling unit consults the relevant DB unit for the
identification of the best VMs (Virtual Machines) for every request. The most attractive VM
is one with sufficient assets to match the consumer criteria and the lowest intake in intensity
and the highest response time.

III. CONCLUSION

Cloud computing has emerged as the go-to solution for on-demand applications, service
aggregation, and computing technology around the world. Many cloud infrastructure
organisations have efficient implementations, improvements in cloud creation, and valuable
great services. When the use of cloud computing grows, so does its effect on the climate as
the need for electricity grows. Everywhere, including cloud computing, increased energy
demand has become a challenge. There are numerous algorithms for a variety of
specifications, many of which perform well in a cloud setting. In general, the search
mechanism in the cloud setting is dynamic, requiring good optimization results for further
maintenance. Cisco estimates that by 2021, cloud infrastructure structures will account for
about 94 percent of computing. Furthermore, the International Data Company (IDC) predicts
that by 2025, the amount of data generated and exploited would have grown to 175 zetta
bytes. This necessitates cloud providers installing additional data centres and facilities.
Green Cloud computing can be an environmentally sustainable model because it protects the
world from being harmed by cloud computing and promotes sustainability. It can also reduce
CPU load and reuse resources, allowing for more consistent and dependable resource usage.
As a result, an environmentally friendly cloud data centre infrastructure is needed. The
architecture should consider the purpose of the cloud service request as well as the
geographic location of the cloud data centre, and it should also address the power
consumption problem.

IV. REFERENCES

[1] Ebtesam H Alharbi1, Maryam M. Alahrbi2, Sahar S. Alkhamali, A PROPOSED FRAMEWORK FOR
ADOPTION GREEN CLOUD COMPUTING IN SAUDI ARABIA, Department of Information Systems.
[2] S. Paul, Green cloud computing for environmental sustainability in UAE, 2018 Adv. Sci. Eng. Technol. Int.
Conf. ASET 2018, pp. 1–4, 2018.
[3] S. Garg, R. Buyya, NetworkCloudSim: Modelling parallel applications in cloud simulations,Proc.2011 4th
IEEE Int. Conf. Util. Cloud Comput. UCC 2011, no. Vm, pp. 105–113, 2011.
[4] Mohammad Shojafar, Riccardo Lancellotti, Adaptive Computing-Plus-Communication Optimization
Framework for Multimedia Processing in Cloud Systems, IEEE

2805 LINGUISTICA ANTVERPIENSIA

 LINGUISTICA ANTVERPIENSIA, 2021 Issue-1
www.hivt.be
ISSN: 0304-2294

[5] Zhong Zong, JSTI Group, Nanjing, An Improvement of Task Scheduling Algorithms for Green Cloud
Computing, 15th International Conference on Computer Science & Education (ICCSE 2020) August 18-20,
2020.
[6] Heng Liu, Xiaofen Zhang, Jing Bi, Bi-objective Intelligent Task Scheduling for Green Clouds with Deep
Learning-based Prediction, Dept. of ECE New Jersey Institute of Technology Newark, NJ 07102 USA
[7] ABDULAZIZ ALARIFI, KALKA DUBEY, MOHAMMED AMOON, TORKI ALTAMEEM, FATHI E.
ABD EL-SAMIE, AYMAN ALTAMEEM, S. C. SHARMA, AND AIDA A. NASR, Energy-Efficient Hybrid
Framework for Green Cloud Computing, Received May 22, 2020, accepted June 2, 2020, date of publication
June 12, 2020, date of current version July 2, 2020. Digital Object Identifier 10.1109/ACCESS.2020.3002184
[8] Christos L. Stergiou, Kostas E. Psannis, Thessaloniki, Greece, Yutaka Ishibashi, Green Cloud Communication
System for Big Data Management, The 3rd World Symposium on Communication Engineering, 978-1-7281-
8564-4/20/$31.00 ©2020 IEEE 2020.
[9] Ahmadreza Montazerolghaem, Mohammad Hossein Yaghmaee, Green Cloud Multimedia Networking:
NFV/SDN Based Energy-Efficient Resource Allocation, IEEE TRANSACTIONS ON GREEN
COMMUNICATIONS AND NETWORKING, VOL. 4, NO. 3, SEPTEMBER 2020.
[10] Xiang Sun, Green Cloudlet Network: A Sustainable Platform for Mobile Cloud Computing, IEEE
TRANSACTIONS ON CLOUD COMPUTING, VOL. 8, NO. 1, JANUARY-MARCH 2020.
[11] Tanu Shree, Rajiv Kumar, Nikhil Kumar, Green Computing in Cloud Computing, Department of Computer
Science and Engineering, 2ndInternational Conference on Advances in Computing, Communication Control
and Networking (ICACCCN), 2020.
[12] Surekha B, Vundavilli PR, Parappagoudar MB, Srinath, Design of genetic fuzzy system for forward and
reverse mapping of green sand mould system , AINTERNATIONAL JOURNAL OF CAST METALS
RESEARCH, FEB 2011 10.1179/1743133610Y.0000000002.
[13] Praveen SP, Rao KT, Janakiramaiah, Effective Allocation of Resources and Task Scheduling in Cloud
Environment using Social Group Optimization , B ARABIAN JOURNAL FOR SCIENCE AND
ENGINEERING, AUG 2018 10.1007/s13369-017-2926-z.
[14] Lavanya K, Reddy LSS, Reddy, Distributed Based Serial Regression Multiple Imputation for High
Dimensional Multivariate Data in Multicore Environment of Cloud , BE INTERNATIONAL JOURNAL OF
AMBIENT COMPUTING AND INTELLIGENCE, APR-JUN 2019 10.4018/IJACI.2019040105.
[15] Bashir AK, Arul R, Basheer S, Raja G, Jayaraman R, Qureshi, An optimal multitier resource allocation of
cloud RAN in 5G using machine learning, NMF TRANSACTIONS ON
EMERGINGTELECOMMUNICATIONS TECHNOLOGIES, AUG 201910.1002/ett.3627
[16] Rathod SB, Reddy VK, NDynamic Framework for Secure VM Migration over Cloud Computing
JOURNAL OF INFORMATION PROCESSING SYSTEMS, JUN 2017 10.3745/JIPS.01.0015.
[17] Sastry, J.K.R., Sai Abhigna, K., Samuel, R., Kamesh, D.B.K., Architectural models for fault tolerance within
clouds at infrastructure level 2017 ARPN Journal of Engineering and Applied Sciences, KAPPLIED
ARTIFICIAL INTELLIGENCE 10.1080/08839514.2019.1689714
[18] Dey NS, Gunasekhar T, A Comprehensive Survey of Load Balancing Strategies Using Hadoop Queue
Scheduling and Virtual Machine Migration IEEE ACCESS 2019 10.1109/ACCESS.2019.2927076.
[19] Ranjeeth Kumar, M., Srinivasu, N., Reddy, L.C., Fine grained multi access control via group sharing in
distributed cloud data 2017 Journal of Theoretical and Applied Information Technology 95 14 3242
3250 https://www.scopus.com/inward/record.url?eid=2-s2.0-
85026759970&partnerID=40&md5=82212e2970dae18155ae7b80257852ad Article Final Scopus 2-
s2.0-85026759970
[20] Jena, S.R., Rao, L.S., A study on energy efficient task scheduler over three-tier cloud architecture using green
cloud, 2017 Journal of Advanced Research in Dynamical and Control Systems, Scopus 2-s2.0-85052077442.
[21] P. Kshirsagar and S. Akojwar, "Classification & Detection of Neurological Disorders using ICA & AR as
Feature Extractor", Int. J. Ser. Eng. Sci. IJSES, vol. 1, no. 1, Jan. 2015.
[22] Pravin Kshirsagar and Sudhir Akojwar, “Hybrid Heuristic Optimization for Benchmark Datasets”,
International Journal of Computer Application (0975-8887),Vol.146- No.7,July 2016 .
[23] ]Pravin Kshirsagar and Dr. Sudhir Akojwar, Novel Approach for Classification and Prediction of Non Linear
Chaotic Databases, International Conference on Electrical, Electronics, and Optimization Techniques, March
2016.
[24] Pravin Kshirsagar and Dr. Sudhir Akojwar, Prediction of Neurological Disorders using Optimized Neural
Network, In the proceeding of International Conference on signal processing, Communication, Power and
Embedded System ,October (2016).
[25] Sudhir Akojwar, Pravin Kshirsagar, “A Novel Probabilistic-PSO Based Learning Algorithm for Optimization
of Neural Networks for Benchmark Problems”, WSEAS International conference on Neural Network-2016,
Rome, Italy.
[26] [26]Sudhir Akojwar, Pravin Kshirsagar, “Performance Evolution of Optimization Techniques for
Mathematical Benchmark Functions”, WSEAS International conference on Neural Network-2016, Rome,Italy.
[27] [27]Pravin Kshirsagar, Dr. Sudhir Akojwar, “Classification and Prediction of Epilepsy using FFBPNN with
PSO”, IEEE International Conference on Communication Networks, 2015.

2806 LINGUISTICA ANTVERPIENSIA

View publication stats