2014 14th International Conference on Computational Science and Its Applications

A Cost Modelling System for Cloud Computing

Daniel Edache Ajeh#1, Jeremy Ellman*2, Shelagh Keogh#3
#
Faculty of Engineering and Environment, Northumbria University,
Newcastle United Kingdom
#1
[email protected], [email protected]
*2
[email protected]
#3
[email protected]

Abstract— An advance in technology unlocks new opportunities production without increase in cost or to cut down cost of
for organizations to increase their productivity, efficiency and production [17]. In recent time, cloud computing has been
process automation while reducing the cost of doing business as marketed as a technology which brings about cost savings,
well. The emergence of cloud computing addresses these
scalability, flexibility and reliability of services. It averts the
prospects through the provision of agile systems that are
scalable, flexible and reliable as well as cost effective. Cloud need for underutilized IT infrastructure in anticipation of peak
computing has made hosting and deployment of computing demands as with very short notice, services can be scaled up.
resources cheaper and easier with no up-front charges but pay Thus, lead to reduced cost of purchasing infrastructure,
per-use flexible payment methods. However, there is lack of tools upgrades, and maintenance as well as energy savings [38] [22].
to aid decision makers in evaluating the much promised benefits These promised benefits have stirred large organizations’
of cloud computing particularly its cost benefit. To fill this gap in interest in beginning to consider cloud computing options.
tools for evaluating the cost benefit of cloud services as an However, there is much doubt and uncertainty in the minds of
alternative to on premise computing, a cost modelling system for decision makers as regards the actual realization of the
cloud computing (CCMS) is proposed. A prototype model was
developed to simulate the cost incurred on maintaining an on
promised benefits, due to much hype and assumptions,
premise IT infrastructure under various usage patterns with the particularly in the cost benefit analysis of adopting cloud
purpose of determining the cost benefit of cloud alternatives. computing services [15].
CCMS assists decision makers with insights on cost savings of Also, there is growing concerns that high energy cost of
adopting cloud alternatives and also demonstrate how the running private data centres may be worsen by government
utilization capacity and cost of acquisition of an infrastructure led carbon taxes [33]. There are also predictions that by 2015,
can influence the cost savings from cloud alternatives. the initial capital cost of IT infrastructure could be exceeded
by the operational cost over a lifecycle of 5-years [3] [4].This
Keywords— Web application, cloud computing, cloud costing, means that more organizations are likely to consider cloud
cost model, cloud analyst, cloud simulator, utilization model, options. Furthermore, many organizations’ drive for change at
cloud modelling. the moment is predominantly viewed from the cost
perspective as they continue to discover how underutilized
I. INTRODUCTION their substantial capital investment into IT is on the increase.
The history of information technology (IT) can be thought It has been noted as well that close to half the capital
of as a pendulum that swings between two extremes; equipment budget goes into IT but the capacity utilization of
centralized and distributed computing. The early mainframes servers is only 6% on the average [39] [38]. Therefore, the
are clear examples of centralized computing whereas the later proposed cost modelling system is a tool that will provides
days of PCs and workstations are a big contrast; with cost benefit analysis of migrating whole or part of the
distributed computing [24]. The internet technology, organization’s IT systems to the cloud and help decision
especially with the improved performance in bandwidth and makers in making more informed decisions regarding IT
connectivity, is taking IT back to a much greater centralized infrastructure procurements and maintenance and in verifying
computing called cloud computing. Among the hottest topics the claims of cloud service providers.
that have emerged in the field of information technology is
cloud computing. It is rooted on more than a few other II. CLOUD COMPUTING ORIGIN
computing technologies such as virtualization, high- Today’s Information Technology facilities are becoming
performance computing (HPC), grid computing and utility more and more complex and expensive. The cost implication
computing [11] [20] [22]. Cloud computing is not only of maintaining data centres which includes capital and
perceived as a buzz in today’s world of computing but a big operational costs of such facilities have direct impact on the
and developed branch of information technology. As proposed profitability of the business processes being driven by them
by a research firm IDC, cloud computing was expected to hit [12]. Historically, what is today known as cloud computing
$42 billion last year [29]. evolved from utility computing and grid computing. Utility
An Organization’s decision to invest in Information computing emerged from the materialization of virtualized
technology (IT) may be for several reasons; some of these are systems for servers, storage and networks which provided
demands to improve quality of product or service, to increase organizations with the pay-per-use or pay-as-you-go services

978-1-4799-4264-0/14 $31.00 © 2014 IEEE 74

DOI 10.1109/ICCSA.2014.24
Authorized licensed use limited to: LIVERPOOL JOHN MOORES UNIVERSITY. Downloaded on February 07,2024 at 00:57:25 UTC from IEEE Xplore. Restrictions apply.
 like that of public utility – a key benefit of this lies in capital capabilities such as hardware, software and services can be
and operational cost savings while a computing environment provided to an organization dynamically with flexibility to
where the workloads are shared or allocated to nodes which scale up or down as the need arises. According to Miller [24],
have necessary computing resources is referred to as grid the 24/7 reliability and universal access coupled with the
computing. Usually, in grid computing, a chain of clustered ubiquitous collaboration guaranteed by cloud computing is
servers are made available to cater for distributed workloads. likely to do away with today’s desktop-centric computing
It is also capable of parallel computing [9]. Cloud computing notion. The way of the future can be seen in cloud computing
can be seen as the computing equivalent of the last century’s [24]. Presently, cloud deployment can be broadly classified
electricity revolution- whereby everyone generated their own into three models, namely; public – where deployed cloud
electricity from single units of generators. When the electrical computing resources are made available for use by the general
grid became operational, everyone gradually powered down public; private – where the cloud computing infrastructures
their generators and got connected to the grid for more are controlled exclusively by an organization and hybrid cloud
reliability at a much lower price. Cloud computing remains a – which is a combination of private and public cloud
rather amorphous term but one that definitely has gained wide deployment model. Private clouds have shown to be more
usage. It is a model that enables network access to a shared secured than public cloud [32] [16].
pool of configurable computing resources in a convenient,
ubiquitous, on-demand manner. With cloud computing, client B. Characteristics of cloud computing
computers only serve to transmit instructions and receive 1) High flexibility: Computing capabilities in the cloud is
results from the remote systems where the computations are rapidly elastic which allows for resources to be quickly scaled
carried out. The users are at liberty to use any computer up or released based on user demands. Cloud consumers often
provided it has connection to the internet. Prospects for experience resource capacities that seem to be unlimited no
improving IT efficiency and performance through centralism matter the demand at any given moment [21].
of resources have increased radically in the last few years with 2) Resource Sharing: The computing resources from the
the development of technologies such as service-oriented providers are pooled to service multiple clients by means of a
architecture, management automation, virtualization, and grid multi-tenant model which dynamically allocates and
computing. Today, what is referred to as cloud computing is a reallocates different resources to users based on demand. For
natural outcome of these developments - where a user of instance, computing resources such as processing and storage
computational capabilities sets up or makes use of computing are often made available to consumers with no control or
in the cloud over a network in a self-service manner, without ability to specify the exact location of such resources [21].
direct involvement in how that computing is resourced [18] 3) On-demand Self-service: Client’s request for computing
[36]. Cloud computing is a new computing paradigm which resources is automated and as such requires no human
pools diverse client devices with computational and data interaction from the service providers. This means a user can
storage capacities to the cloud. The emergence of cloud gain access to computing capabilities unilaterally [21].
computing as one of the current topics in the field of 4) Extensive network access: The cloud resource
information technology calls for a proper understanding of the capabilities are available to consumers via standard
domain and how individuals, private and public entities can mechanisms which support heterogeneous client platforms
leverage on its much promised benefits [1] [11]. Cloud such as workstations, laptops, tablets or mobile phones [21].
computing paradigm may be traced back to the early 60s but 5) Measured Service: The cloud computing capabilities are
since then there has been no commonly accepted definition for provided to users through metering systems that automatically
it until September, 2011 when the United States National monitor, control and report resource usage to both the users
Institute of Standards and Technology (NIST) released the and the providers. Hence, making the whole process
16th and final version of its definition [21]. transparent to the duo involved [21].
6) Pay Per-usage: Cloud resources are provided to
A. What is cloud computing? consumers on pay per-usage. This eliminates up-front charges
Cloud computing simply means the ability to access and and allows users to release resources at the earliest time when
utilize computing resources such as storage, applications and they are not needed. In fact, cloud users see the utility based
processing power via internet [31]. Although there have been payment method of the cloud as a welcome development and
several definitions of cloud computing, the academic considers it as a main gain of cloud computing [29].
definition was first given by Kenneth K. Chellappa as "a
computing paradigm where the boundaries of computing will C. The Benefits of Cloud Computing
be determined by economic rationale rather than technical Qaisar [31] argues that even planning well ahead does not
limits" [30]. According to NIST [21], cloud computing is eliminate the two likely outcomes of maintaining an on
defined as “a model for enabling ubiquitous, convenient, on- premise computing model. Maintaining an on premise
demand network access to a shared pool of configurable computing results either on wasted capital in acquiring excess
computing resources (e.g., networks, servers, storage, infrastructure capacity or constrained capacity due to limited
applications, and services) that can be rapidly provisioned and infrastructure. Such resources could be better put to use for
released with minimal management effort or service provider other strategic plans. Apart from the high capital cost involved,
interaction ”. This means that with cloud computing IT the ever changing demand and complexity of configuration

75

Authorized licensed use limited to: LIVERPOOL JOHN MOORES UNIVERSITY. Downloaded on February 07,2024 at 00:57:25 UTC from IEEE Xplore. Restrictions apply.
 management, regular patching and required upgrades is not computing is less energy efficient than cloud computing. It
minor and this often stall the agility of an organization [31]. It has been argued that offloading of computation and storage
is wise to think of how long it takes to acquire and setup a requirements to the cloud will yield better energy efficiency.
new infrastructure or system and have it ready for use; Also another research group [2] perceives that cloud
perhaps the idea of offloading part of an organization's computing seems an alternative to desktop computing in terms
computing needs to the cloud would be worthwhile. Cloud of energy consumption. At an Uptime Institute Symposium
computing demand is driven by innovation, consumer demand held in New York Jonathan Koomey, a data centre energy
and rising devices. This means that there will be a reliable expert, recommended cloud computing as an energy saving
infrastructure and network environment and abundance of tool [25]. In 2010, Nucleus Research in a report says that
access to network services. It lowers start-up cost, speeds up businesses running the on premise desktop computing use
deployment and is scalable. It has been argued that it gives 91% more energy than business users utilizing cloud
access to very high performance data centre space in the world. computing [26]. These reports from various researchers above
It allows locations of assets and applications in proximity to leaves one in no doubt that the decision to partially or fully
the users to improve performance. It provides easy access to migrate an organization's computing needs to the cloud will
the widest variety of services, platforms and applications on significantly lower energy consumption which also means
demand with flexibility and easy scalability. Computing in the lower cost.
cloud provides acceleration of application performance which
improves the end users experience and return on investment. D. The hidden requirements of cloud computing
In today's digitalized world, no matter what service or product 1) Cloud Implications: An implementation of cloud
being offered, a strategic differentiator for business is computing will mean a major change in the way information
performance. It impacts on business and this translates to will be stored and how applications will be run. Unlike
either increase or reduction in revenue. Cloud services are running programs and data from individual computer systems,
designed for rapid deployments and easy adaption to changing everything is hosted and accessed from the cloud. This comes
business needs. It minimises upfront capital investments and along with such a huge benefit of allowing access to all
rip benefits of more flexible operating expenses. applications and documents from anywhere anytime around
Computing in cloud eliminates the need for user’s up-front the world. This eliminates the limitation of users being
capital commitment to computing resources, thereby allowing confined to their desktops and provides an easy means for
organizations or individuals to start small and scale up collaboration among team members. More so, the ubiquitous
resources only when the need arises. Also the provision of pay online presence so demanded by modern businesses and
per use which allow users to pay and utilize computing projects would be greatly achieved while providing
resources on short-term basis as needed and release same opportunity for significant cost saving, internet based access,
resources when not in use amounts to so much cost savings as workload balancing, unlimited scalability, dynamic and
computing resources are not tied down when they are not granular allocation of resources with self-servicing request to
utilized. Reduces cost through elimination of expensive IT users [36].
infrastructure and specialized staff to deploy, operate and 2) Consultancy: In a recent Cloud Computing event
upgrade systems. It swiftly accommodates business growth held in London [7], it was made very clear that the journey to
through scalability solutions. Researchers have argued that the the cloud could include some hidden cost. Such cost
power consumption of modern computer systems is not as highlighted by one of the cloud service providers SAP, was
efficient as that of cloud software service [28]. that of consultancy. This cost is said to be varied as this
1) Pay Per-usage: Pay per usage is a very key feature of depends on several factors like the size of the organization,
cloud computing. It permits users of cloud services to request the volume of the data to be migrated, the number of users and
and utilize only necessary resources needed and made the volume of transactions within the organization [7].
available by cloud service providers for a specified period at a 3) Resource Usage Capacity. Performance monitoring
given cost. This eliminates up-front charges and allows users and resource usage are critical on the journey to the cloud.
to release resources at the earliest time when they are not The average usage capacity of the computing resources needs
needed. In fact, cloud users see the utility based payment to be known in order to estimate the cost of acquiring same in
method of the cloud as a welcome development and considers the cloud [7]. This demands very good monitoring tools to be
it as a main gain of cloud computing [29]. put in place for monitoring, recording and reporting resource
2) Energy efficiency of cloud computing: There has utilization over a given period. Such period will include peak
been a growing concern about the increased rate of carbon periods and off-peak periods. This does not only mean
emission from the activities of ICT due to the expansion of the additional cost of acquiring such tool but enough planning
information community and the introduction of new devices time to enable a smooth and successful migration into the
and services [10]. A huge part of the cost of running the cloud.
traditional on premise computing can be attributed to the 4) Internet Services: The whole concept of cloud
energy cost which does not only includes the cost of powering computing relies solely on the internet for connectivity
the IT infrastructure but also the cost of cooling the equipment. between the cloud user and the provider. This fundamental
Some researchers have proposed that the traditional desktop requirement is usually silent during discussions on migration
to the cloud. According to some of the cloud service providers

76

Authorized licensed use limited to: LIVERPOOL JOHN MOORES UNIVERSITY. Downloaded on February 07,2024 at 00:57:25 UTC from IEEE Xplore. Restrictions apply.
 at the Cloud Computing World Forum event, readiness for Although the rent rate differs from place to place, Officegenie
migration to the cloud means that internet connectivity and [27] suggests that the space required for a system setup is
bandwidth are no issues [7]. In fact, some providers say this about 23 square meters. This will cost about £90.00 in the UK
was the reason why their services are limited to only the depending on the location.
developed countries where persistent internet connectivity is
not an issue. This then means that organizations located within A. The Financial Model
developing countries where problems of reliable and The financial model analyses the various cost components
persistent internet connectivity and bandwidth exist are not of an infrastructure and presents the cost benefit analysis of
only likely to lose some of the benefits of cloud computing considering available cloud options. The financial model
after migration but will also need to include the cost of considers basic cost components involved in acquiring and
internet connectivity in their budget when planning for maintaining an information technology infrastructure. These
migration to the cloud. Furthermore, researchers [28] have include the cost of purchasing new systems (Cpc), the cost of
suggested that the shift from desktop computing to the cloud rent on the space where the computer systems are setup for
will mean increasing demand on the communication networks use (Cs), the cost of setup and configuration of the computer
in terms of support and energy consumption. systems (Csc), the cost of power consumption (Cp), the cost
of technical support and maintenance of the systems (Cm) and
III. COST MODELLING CONCEPTS the cost of depreciation on the infrastructure (Cd). The
Cost accounting has been described as a discipline following scenario is used just for the purpose of the model
embraced by decision makers in order to plan, make decisions development.
and control the cost of cost objects such as products or Adoka University have several computer Labs, each
services. The chief aim of cost accounting is modelling cost dedicated to training and research studies. Students of various
objects as accurately as is economically reasonable [23]. In courses in the School of Engineering and Computing Science
order to make good business decisions, accurate cost make use of the labs at different times during the course of
accounting is crucial because it confirms the efficient their training. These Labs are equipped with the latest
management of resources and help decision makers in making computer systems, powered and well maintained to ensure
the right choice when faced with multiple alternatives of minimal downtime, high efficiency and availability
investment. Mikko et al [23] further argued that determination throughout the year. The systems in each of the labs are
of advantageous prices for a particular product or service can upgraded to new ones after every five (5) years. Although the
be provided with the help of cost accounting data. This labs are available for use throughout the year, the systems are
ensures wise spending since such data reveals the total cost of mostly utilized only within the official hours. The chief
ownership of a particular product or service. In the case of IT information officer is considering subscription to cloud
infrastructures, these will include the cost of acquisition, services for one of the labs but needs to justify his proposal
configuration, installation, licensing and management of the with cost benefit analysis before the school management
given product or service. It also brings about efficient unit board.
costing that help to justify IT resource [23]. These cost Considering one of the labs equipped with thirty (30)
incurred may further be broken down into cost of hardware, systems at the cost of £600.00 each with no salvage value
software, space, labour, power and outsourcers [8] [14]. after an economic life of five (5) years. Assuming the rent for
However, cost can be broadly classified into two types, the lab space is at the current market price, the maximum
fixed cost and variable cost. Cost that remain constant power consumption per system is 180w per hour (0.18 KWh)
regardless of the level of production or activity are referred to and the cost of maintenance for each system is at the rate of
as fixed costs (Cf) whereas costs that varies in proportion to £100.00 per month. The cost of setup and configuration is
the level of production or activities are referred to as variable £120.00 per system and the power utility charge is £0.15 per
cost (Cv) [35][20]. Relating this to acquisition and KWh [19][34]. Assuming the power utilization efficiency
maintenance of IT infrastructures, some examples of fixed (PUE) is 1.7, the financial analytical model is shown below.
costs will include cost of hardware, software, licenses and Analytical Calculation
location space while the cost power supply, cooling, and The total number of instance (N) = 30
maintenance support are examples of variable cost. Hence, the The cost of system acquisition (Cpc) = 1000
cost of maintaining on premise computer systems is the The cost of power = (Cp)
accumulation of these cost components involved in The cost of space = (Cs)
purchasing and maintenance [8] [20]. According to Grisebach The cost of maintenance = (Cm)
[13] a typical computer system can consumes up to 175 watts The cost of depreciation = (Cd)
of power (0.18 KWh) excluding the power consumption of the The cost of setup and configuration = (Csc)
monitor. The charge for every unit of power consumed is The salvage value = (SV)
about £0.15 per KWh [34]. This has a significant effect on the Economic Life = (EL)
overall running cost of the systems over a specified period of Total cost of ownership of infrastructure = (TCO)
time. The space requirements for setup also affect the cost as Cost of depreciation (Cd)
the size of space will determine the amount paid for rent. The annual depreciation on the systems is determined by
the economic life of the system, the salvage value and the cost

77

Authorized licensed use limited to: LIVERPOOL JOHN MOORES UNIVERSITY. Downloaded on February 07,2024 at 00:57:25 UTC from IEEE Xplore. Restrictions apply.
 of acquisition of the system. The lower the salvage value, the Total number of hours per year = 8,736 hours
higher the depreciation and vice versa. Total idle time in a year = 8736 - 1200 = 7,536 hours
Annual cost of depreciation (Cd) = (Cpc - SV)/EL The % utilization in hours = (1200/8736 x 100) = 13.7 %
Cd= (600 -0)/5 =£120 per system The % idle time in hours = (7536/8736 x 100) = 86.3 %
Cost of power consumption
The total cost of power consumption by the computer B. The CCMS Model Architecture
systems is determined by the power consumption rate per The cloud computing modelling system (CCMS) is a decision
system per hour multiplied by the utility charge and the power support tool designed to compute, analyse and present useful
utilization efficiency (PUE). The power utilization efficiency analytical reports on an infrastructure cost, utilization gradient
is a metric used to express the relationship the ratio of the and the cost benefits of using available cloud option for same
power consumed by a datacentre infrastructure to the power organization’s Information Technology infrastructure
supplied to the datacentre. This relationship is established as a computing needs.
numerical value that can be used as basis for understanding Register Cloud service providers
the efficiency of the energy consumption of a datacentre. (Administrator only)
Cost of power per hour (Cp) =0.18 x 0.15 x 1.7 = 0.05 Select System Specification (Small, Medium,
Total number of hours per year = 8 x 5 x 30 = 1200 Large)
Setup/Start-up Specify current infrastructure cost parameters
Annual cost of power per system = 0.05 x 1200 = 55.08 Phase Specify usage pattern (hour, week, and year)
Cost of Maintenance
The annual cost of technical support and maintenance (Cm)
on one system is given by the monthly cost of multiplied by
12 months.
Cm = 100 x 12 =£1,200.00
Cost of Setup and Configuration
The cost of setup, configuration and testing is a one off cost
incurred when setting up new systems (Csc) =£120.00 per Compute cost of rent, cost of power, cost of
system depreciation, annual total cost of ownership
Cost of space Computation
(TCO) of current infrastructure, hourly TCO and
The cost of space varies according to the location of an hourly TCO per system
Phase
office. The cost is usually calculated based on the square feet.
According to a property management company based in the
United Kingdom [27], the average space required per system
is about 23 square meters which cost about £4.00 per square
meter per month.
The annual cost of space (Cs) for 1 system = 4 x 12 x 23 Analyse percentage utilization, percentage idle
=£1104.00 time, the breakeven point of current infrastructure
Estimated Annual Total Cost of Ownership of Analytic in the cloud and the cost savings.
Infrastructure (TCO) Phase
This is the sum of the fixed and variable cost incurred on
systems for the period of one year.
TCO= {[(Cpc+ Cp+ Cm+ Csc+ Cd+ Cs) x (N)]}
= {[(600.00+ 55.08+ 1200.00+ 120.00+ 1104.00) x (30)]} Display analytical reports and recommendation
= £95,972.40 per year Reporting based on the cost savings .
Phase
Hourly Cost of Ownership
The hourly cost of ownership = (TCO/Hours)per year
(where hours represents the usage hours) = 95972.40/1200
= £79.97 per hour Figure 2 CCMS Architecture
Hourly cost per system
The hourly cost of ownership for one system cost The design of the CCMS model is produced considering
= Hourly TCO/N three business scenarios. It shows how the system functions
=£2.66 per hour and defines the exact stages of operation of the system. The
Actual Utilization Hours model architecture is broken down into four stages; the start-
The systems are actively used during working hours which up phase, the computation phase, the analytic phase and the
is between 9am and 5pm for only five days a week. This reporting phase. Each of these stages has unique set of
amounts to a total usage hour of 40 hours per week. The functionalities performed to achieve the overall goal of the
number of weeks utilized by the school sessions in a year is model which is to present to the user the cost benefit of
thirty (30) weeks. This means that the actual utilization hours considering cloud options as against their current
in a year = 40 x 30 = 1200 hours infrastructure. In figure 1 above the CCMS model is clearly
defined.

78

Authorized licensed use limited to: LIVERPOOL JOHN MOORES UNIVERSITY. Downloaded on February 07,2024 at 00:57:25 UTC from IEEE Xplore. Restrictions apply.
 powered off, the breakeven point of the current infrastructure
CCMS MODEL
in the cloud and the cost savings from cloud options.

IV. MODEL IMPLEMENTATION

Select system
specification
Register cloud
providers
The CCMS model is implemented as a web application
*
** * * **
*
using HTML, CSS, PHP, MySQL, and Javascripts. The
Specify cost
choice of implementing the model as an online web
User *
components
Register system Administrator application is because of the many advantages of web
specification * applications such as universal access to users across
geographically dispersed locations, fast and easy updates to
View analyses and
*
end users when new versions are released. Online application
* recommendation
Set providers cost
for specifications
also means that there is no need of installation of the
application on individual client systems as it is hosted
centrally on a web server and allows users to have access to
the application irrespective of their operating systems because
the application is platform independent. Also, interested users
can even analyse their IT infrastructure from their mobile
Figure 2 Use case showing the functional requirements of the CCMS devices provided they can browse the internet. Furthermore,
Model there is no need of downloading and installing any software
which additionally saves time and cost for the application user.
1) Start-up Phase The CCMS model has seven different screens beginning with
This is the stage when the CCMS set system ready to perform the login screen and ending with the recommendation screen.
the required analysis. At this stage the administrator will first The login screen is designed to allow new users to sign up and
register the cloud service providers and the pricing of their existing users to login and access the system. The system
services according to system specifications after which the registers the analysis of the user and this can be tracked by the
user login to the system, selects a system specification and administrator from the admin console.
supply all the necessary parameters data according to their Figure 3 below shows the login screen.
current infrastructure. The system proceeds to the next stage
as soon as the user completes the last step of this stage.
2) Computation Phase
The computation phase is entirely handled without the user’s
interaction. The system compute the estimated cost of the
user’s current infrastructure based on the input parameter
values. At this stage the estimated total cost of ownership is
computed based on the usage pattern. The usage hours is also
used to estimate the percentage utilization of the infrastructure
and the percentage idle time. The model further establishes a
breakeven point of the current infrastructure on the cloud. It
also computes the cost saving for an equivalent infrastructure
from any of the available cloud options. Figure 3 showing login screen of CCMS model
3) Reporting Phase The next screen to the login is that for the system
This stage of the system presents the output of the whole specification. It enables the user to select from the list of
analysis to the user on the screen. The results presented here categories (small, medium, large) that corresponds to the
are modular and they are displayed to the user on different intended infrastructure on the cloud. Once a selection is made,
screens, each allowing for easy understanding. At this stage the ‘proceed’ button takes the user to the next screen where
the user is also presented with a recommendation screen the user can estimate the annual costs on current infrastructure.
showing the preferred solution based on cost savings. This screen is well enhanced with sliders using Javascript
Additionally, the administrator can also track the various users language to enable easy and automatic manipulation of the
that have successfully analysed their infrastructure on the values by adjusting the slider corresponding to each cost
system. component. The next screen presents the user with the current
4) Analytic Phase infrastructure utilization analysis. The information shown on
The analytic phase takes place after all the start-up and this screen includes the annual idle time, annual utilization
computation phases are completed. This phase is responsible time, percentage utilization and percentage idle time. The
for investigating the results of all computation derived from ‘proceed’ button from the utilization analysis screen takes the
the input parameter specified at the star-up phase. The result user to the breakeven screen. This screen presents the user
of the analysis at this stage include the percentage utilization with the view of what the cost of one system in the current
of the current infrastructure based on the active hours of the infrastructure can acquire on either Amazon cloud or
system, the percentage idle time when the systems remains Microsoft cloud platform.

79

Authorized licensed use limited to: LIVERPOOL JOHN MOORES UNIVERSITY. Downloaded on February 07,2024 at 00:57:25 UTC from IEEE Xplore. Restrictions apply.
 The cost benefit analysis screen is the next screen to the Simulation Results
breakeven screen. This screen presents the user the cost Scenario School Lab PC
savings from each of the cloud providers. It compares the cost
Number of systems 30
for equal number of systems and usage hours and shows the
cost savings for each provider. Figure 10 above shows cost Cost of system 600
benefit analysis of an infrastructure. The last screen is the Usage hours 1200
recommendation screen which presents to the user the annual Percentage utilization 13.74%
TCO from each provider including the current infrastructure Percentage idle time 86.26%
and makes e recommendation based on the lowest cost value. Total cost of ownership of infrastructure
In order to review previous screens, the user may click on the (TCO) 95,972.40
back button to review the previous output. The screen for the Hourly cost of ownership of infrastructure
annual cost of current infrastructure can only be reviewed by (TCO) 79.98
setting the sliders to the position that represents the correct Hourly cost of ownership of one system 2.67
value for each cost component. On the recommendation Amazon cost 0.07
screen, the user can choose to exit the application by clicking Microsoft cost 0.05
on the logout button. Break even for one system on Amazon 36
V. MODEL ANALYSIS AND EVALUATION Break even for one system on Microsoft 53
Cost Savings on Amazon 93,240.00
A. Scenario 1 – University Lab Cost Savings on Microsoft 93,960.00
This section considers the usage scenario of systems in a
school lab to analyse their usage pattern and cost of ownership Table 1 showing the simulation results for a school lab
of the systems as compared to the cost of owning same in the infrastructure
cloud. This scenario is used to gain an understanding of the
percentage utilization of an infrastructure particularly in a
school lab. The test case was simulated with parameter data of
thirty (30) systems setup in a lab dedicated to the learning of
system forensics based on the usage pattern of 8 hours a day,
5 days a week and 30 weeks a year. This accounted for 1200
hours usage in a year out of the total of 8736 hours in a year.
The cost of the systems was placed at £600.00 per system and
the maintenance cost per system at £100.00 per system. The
estimated cost of setting up and configuring each system was
put at £120.00 and the salvage value of the systems was
assumed to be zero after an economic life of 5 years. Other
parameter data considered were utility rate of 0.15
pence/KWh [34] and power consumption rate of each system
at 180 watts or 0.18 KWh. Taking the space requirement of 23
square meters at the cost of £4.00 per square meter, the case
scenario was simulated on the model and the results are shown
in Table 4 below.
Results
From the simulation result of the model, the percentage
utilization of the Lab systems stands at 13.7% with an idle Figure 4 showing the percentage utilization of a school lab
time of 86.3% and this is represented by the chart in figure 4 infrastructure
below. Also from the result of the simulation it was shown Analysis
that the estimated hourly cost of owning such an infrastructure These results can be useful for planning such IT infrastructure
is about £80.00 per hour which means an hourly cost of one to determine from the start whether to use cloud services or on
system will be £2.67 per hour. This amount can be used to premise computing resources. It then means that by choosing
acquire up to 36 systems on an Amazon cloud platform and 53 the cloud computing alternative on Amazon or Microsoft, the
systems on a Microsoft cloud platform. school lab will annually save up to £93,240.00 or £93,960.00
respectively. Also going by the evaluation results obtained
from the simulation, the benefits of cloud adoption in this case
will not be limited to just cost saving from choosing any of the
cloud alternatives but the pay per-usage feature with
flexibility of usage hours, easy access to an elastic platform
that can be scaled up or down at any time with the absence of
upfront payments for the needed computing resource.
Furthermore, a consideration of the cloud alternative will also

80

Authorized licensed use limited to: LIVERPOOL JOHN MOORES UNIVERSITY. Downloaded on February 07,2024 at 00:57:25 UTC from IEEE Xplore. Restrictions apply.
 imply that the economic life of the current systems can be Simulation Results
extended which means more cost saving from purchasing new Scenario Office PC
systems and upgrade processes. 30
Number of systems
B. Scenario 2 – Office usage Cost of system 600

The previous section considered systems in a school lab Usage hours 2080
where the usage pattern accounts for only 30 weeks in a year. Percentage utilization 23.81%
In this section the prototype model is evaluated considering 52 Percentage idle time 76.19%
weeks usage pattern of desktop systems in 30 offices. The Total cost of ownership of infrastructure 97,184.16
systems are powered on 8 hours a day and five days a week. (TCO)
The scenario here is used to provide an understanding of the Hourly cost of ownership of infrastructure 46.72
percentage utilization and idle time of systems used in offices (TCO)
for applications processing and comparison of the cost of Hourly cost of ownership of one system 1.56
ownership with available cloud options. This test case is Amazon cost 0.07
simulated with parameter data of thirty (30) systems setup in Microsoft cost 0.05
30 different offices for standard office application usage Break even for one system on Amazon 21
which include word processing and internet browsing
Break even for one system on Microsoft 31
accounting for 2080 hours of use in a year. Again the cost of
Cost Savings on Amazon 92,976.00
the systems was maintained at £600.00 per system with the
maintenance cost of £100.00 per system. The estimated cost Cost Savings on Microsoft 94,224.00
of setting up and configuring each system was put at £120.00 Table 2 showing the simulation results for Office PC
and the salvage value of the systems was assumed to be zero infrastructure
after an economic life of 5 years. Other parameter data Analysis
considered were utility rate of 0.15 pence/KWh [34] and An implication of the above results is that the cost incurred
power consumption rate of each system at 180 watts or 0.18 on one system will be enough to provide access to 21 systems
KWh. Taking the space requirement of 23 square meters at the on Amazon cloud or 31 systems on Microsoft cloud. This then
cost of £4.00 per square meter, the scenario was simulated on means that by considering migration to the cloud options
the model to yield the results shown in Table 2 below. provided by Amazon or Microsoft, the annual savings from
Results running same business processes in 30 offices can be up to
The usage of the office systems shows that the percentage £92,976.00 or £94,224.00 respectively. The amount saved
utilization of the systems stands at 23.8% with an idle time of could be used to purchase cheaper systems with capabilities of
76.2% and this is represented in a chart in figure 5 below. internet connectivity and which will have longer economic life
Also, the results obtained from the simulation showed that the since their usage is for connectivity only. Furthermore, the
estimated hourly cost of ownership on such infrastructure is results above can also be used for planning the computing
£46.72 per hour. This translates to an hourly cost £1.56 per budget for each employee within their organization as the
system. The same amount can be used to procure as much as system also generates the hourly cost of ownership for each
21 systems on the Amazon cloud platform and 31 systems on system in the organization IT infrastructure.
Microsoft cloud platform.
C. Scenario 3 – Dedicated Servers
This section evaluates the usage scenario of dedicated servers
to analyse their usage pattern and cost of ownership of the
systems as compared to the cost of owning same in the cloud.
The significance of this scenario is to gain a better
understanding of the utilization percentage of such
infrastructure hosted within the data centre of organizations
and assert the cost benefit of its cloud alternative. The test
case was simulated with parameter data of thirty (30) Servers
setup in data centres across thirty (30) branch offices for
dedicated online services to users throughout the year. The
servers are meant to provide ubiquitous online access to
different applications. This implies a usage pattern of 24 hours
a day, 7 days a week and 52 weeks a year and accounting for a
total of 8736 hours annually. The cost of each Server system
was placed at £3000.00 with a monthly maintenance cost per
system at £100.00 per system. The estimated cost incurred for
Figure 5 showing the percentage utilization of office systems setting up and configuring each server was £120.00 and at the
end of an economic life of 5 years the salvage value of the
systems was assumed to be zero. Some other parameter data

81

Authorized licensed use limited to: LIVERPOOL JOHN MOORES UNIVERSITY. Downloaded on February 07,2024 at 00:57:25 UTC from IEEE Xplore. Restrictions apply.
 considered were utility rate of 0.15 pence/KWh [34] and opened or closed down. Furthermore, a consideration of the
power consumption rate of each system at 180 watts or 0.18 cloud alternative will also imply that the organization can be
KWh. Also, assuming the space requirement of 23 square free of frequent upgrade and complex configurations due to
meters at the cost of £4.00 per square meter, the model was changing business needs and the results obtained here can also
evaluated by simulating the above data and the results are be used to estimate the losses incurred by an organization in
shown in Table 3 and represented in figure 6 below. an event of a system failure or breakdown.
Simulation Results
D. Summary of Findings
Scenario Dedicated Servers
Number of systems 30 The evaluation results above has demonstrated the capability
3000
of the model as a suitable decision support tool in analysing
Cost of system
the cost benefit of cloud computing. The findings indicates
Usage hours 8736
that the usage pattern and cost of acquisition of systems when
Percentage utilization 100.00% new has a high influence on the utilization capacity, the total
Percentage idle time 0.00% cost of ownership and the cost savings from adopting cloud
Total cost of ownership of infrastructure 192,749.47 alternatives. This implies that certain usage pattern can
(TCO) guarantee better cost savings in the cloud. Also, an extended
Hourly cost of ownership of infrastructure 22.06 economic life of an infrastructure can reduce the total cost of
(TCO)
ownership which means more cost saving.
Hourly cost of ownership of one system 0.74
Amazon cost 0.07 E. Related Work
Microsoft cost 0.05 Recent academic research reveals growing interest in the
Break even for one system on Amazon 10 challenges faced by enterprises in cloud adoption. Presently,
15 matured toolkits or techniques are not yet available to support
Break even for one system on Microsoft
decision makers. Top management of enterprises such as the
Cost Savings on Amazon 175,593.60
Chief Information Officers and Information Technology
Cost Savings on Microsoft 180,835.20 Managers face serious challenges when they need to take
Table 3 showing the simulation results for dedicated servers’ decisions concerning cloud adoption due to lack of decision
infrastructure support tools [15]. In a recent event [7] efforts are being made
in this regards to bridge the gap through consultancy services
provided by several companies such as Accenture, AsterCloud
and AppDirect. The limitation of this kind of approach can be
seen in two ways: such solutions apart from not being
universally available are based on proprietary tools; and often
involve high cost of consultancy periods. In contrast, given
the Cloud Computing Modelling System (CCMS), enterprises
can do without outside consultancy and easily analyse the
economic sense of cloud computing to their organization as
compared to continuing in the traditional or on premise
computing. The CCMS might also be used to verify some of
the claims made by cloud service providers and IT consultant
about the cost benefits of the cloud.
Cloud computing have witness developments of decision
support tools in recent time. The Buyya's lab recently released
a toolkit called CloudSim which serves as a valuable toolkit
for developers interested in modelling and simulating the
cloud computing environment [5] [6]. This toolkit has been
suggested as a useful tool to programmers interested in
Figure 6 showing the percentage utilization of dedicated modelling the performance of their applications in the cloud
servers and also to cloud service providers who are concerned about
Analysis the properties and resource usage of data centres. Researchers
From the evaluation results above it then means that by [37] have also developed a CloudAnalyst, which is a virtual
choosing a cloud alternative on Amazon or Microsoft, an modelling tool for analysing cloud computing environment
organization with such infrastructure can annually save up to and applications. According to the group, the tool is meant to
£175, 593.60 or £180,835.20 respectively. Also the evaluation bridge the gap created due to lack of evaluation tools for
results obtained from the simulation shows that in an event of developers who are interested in evaluating computing servers
server breakdown the cloud can be a smart option for more and user workloads requirements of geographically distributed
cost savings with the pay per-usage feature and it also allow large scale cloud applications. Similarly, a Cloud Adoption
for easy scale up or scale down when branch offices are Toolkit have also been developed to fill the gap of existing

82

Authorized licensed use limited to: LIVERPOOL JOHN MOORES UNIVERSITY. Downloaded on February 07,2024 at 00:57:25 UTC from IEEE Xplore. Restrictions apply.
 research in supporting decision making for deployment of on premise computing infrastructure as compared to the cost
complex IT systems in the cloud. Additional features of the benefit of cloud computing. Therefore, the developed
toolkit also suggest utilization patterns as well as different prototype model of CCMS will evolve over time and result in
pricing schemes of cloud service providers [15]. In contrast an improved quality tool in terms of the analysis it supports.
the Cloud Computing Modelling System simplifies the The model has the potential of being extended to generate
migration journey to the cloud by not just supporting decision graphical output of simulated results of analysis in the form of
maker in the cost benefit analysis of considering cloud options charts. This will aid users to identify important patterns of the
but goes further to break down the journey into stages of output results. Also, a registered user may be granted access to
simple tasks: consideration of system specification; view records of previous analysis without having to start
consideration of fixed and variable cost incurred on afresh. Furthermore, the present model may also be extended
maintaining current IT infrastructure; specification of the to enable users to simulate and analyse mixed usage scenarios
usage hours of the infrastructure, analysis of the percentage with mixed system specifications.
utilization of the current infrastructure and comparison of the
current infrastructure cost of ownership against available REFERENCES
cloud options from renown cloud service providers. [1] A. E. Youssef and M. Alageel, “A Framework for Secure Cloud
Computing”, International Journal of Computer Science Issues (IJCSI),
VI. CONCLUSION Vol. 9, Issue 4, No 3,pp. 478-500, July 2012.
[2] Baliga, J., Ayre, R.W. A., Hinton, K. & Tucker, R.S. “Green Cloud
With organizations faced with several decision making Computing: Balancing Energy in Processing, Storage, and Transport”,
challenges to meet their dynamic business environment due to Proceedings of IEEE, 2011, Vol. 99 (1), pp. 149 – 167.
[3] Barroso, L.A. & Hölzle, U. The Datacenter as a Computer: An
changes in business processes, legacy systems, outdated IT Introduction to the Design of Warehouse-Scale Machines, California:
systems, high operational cost and lack of wider data access, Morgan & Claypool, 2009.
there is a new gap for tools to study and evaluate the benefits [4] Brown, D.J. & Reams, C. “Toward Energy-Efficient Computing”,
of technological solutions and the best time to apply them. Communications of the ACM, 2010, Vol. 53 (3), pp. 50 – 58.
[5] Buyya, R., Calheiros, R.N., Ranjan, R. & De rose, C.A.F. (2009)
The Cloud Computing Modelling System (CCMS) is a new ‘CloudSim: A Novel Framework for Modeling and Simulation of
tool developed to address this gap. It is not just a handy Cloud Computing Infrastructures and Services’ [Online]. Available:
decision support tool for decision makers looking at http://arxiv.org/ftp/arxiv/papers/0903/0903.2525.pdf
evaluating the cost benefit of cloud computing but will also [6] Buyya, R., Ranjan, R. & Calheiros, R.N “Modeling and simulation of
scalable Cloud computing environments and the CloudSim toolkit:
serve as a useful tool for cloud service providers who also Challenges and opportunities”, IEEE International Conference on High
want to verify their claims of lesser cost of computing in the Performance Computing & Simulation, 2009, pp. 1 – 11.
cloud. [7] CCWF (2012) 5th Annual Cloud Computing World Forum, London.
It has been demonstrated that real world scenarios can be [8] D. Gmach, HP Labs., Palo Alto, CA, USA , J. Rolia , L. Cherkasova
“Resource and virtualization costs up in the cloud: Models and design
simulated and analysed using CCMS application to determine choices”, IEEE/IFIP International Conference on Dependable Systems
the percentage utilization of an organization’s infrastructure, & Networks (DSN), 2011, pp. 395 – 402.
the breakeven of same infrastructure in the cloud, the cost of [9] EMC (2010) Cloud Computing Security Overview
the same infrastructure in the cloud and the cost savings from [10] Gartner (2007) 'Gartner Estimates ICT Industry Accounts for 2 Percent
of Global CO2 Emissions' [Online]. Available at:
the available cloud alternative. The finding indicates that the http://www.gartner.com/it/page.jsp?id=503867 [Accessed February 21,
usage pattern and cost of acquisition has a high influence on 2014].
the total cost of ownership, the utilization capacity and the [11] Gong, C., Liu, J., Zhang, Q., Chen, H & Gong, Z, “The Characteristics
cost savings of adopting cloud alternatives. This implies that of Cloud Computing”, Parallel Processing Workshops (ICPPW), 2010,
pp. 275 – 279.
certain usage pattern can guarantee better savings in the cloud. [12] Greenberg, A., Hamilton, J., Maltz, D.A. & Patel, P. (2009) ‘The Cost
Also, an extended economic life of an infrastructure can of a Cloud: Research Problems in Data Centre Networks‘, [Online].
reduce the total cost of ownership which means more cost Available at: http://www.infoworld.com/d/security-central/gartner-
saving. seven-cloudcomputing-security-risks-853?page=0,1 [Accessed
February 21, 2014].
This tool can also be useful for planning an organization’s [13] Grisebach, M. (2003) Information Systems and Technology: Hardware
IT infrastructure in order to determine from the start whether Power Consumption [Online] Available at:
to use cloud services or on premise computing resources. http://windows.uwaterloo.ca/hardware/pc_power_consumption.asp
Management can also use the system to plan the computing [Accessed February 21, 2014].
[14] Jun, T. & Zhongchuan, L. “Study on IT Service Cost's Calculate and
budget for each employee within their organization as the Charging Strategy Based on Activity-based Costing”, IEEE
system also generates the hourly cost of ownership for each International Conference on Service Systems and Service Management,
system held in the organization’s IT infrastructure. The results 2007, pp. 1 – 5.
obtained from the model can also be used to estimate the [15] Khajeh-Hosseini, A., Greenwood, D., Smith, J. W. & Sommerville I.
“The Cloud Adoption Toolkit: Supporting Cloud Adoption Decisions
losses incurred by an organization in an event of a system in the Enterprise”, Journal of Software: Practice and Experience, Vol.
failure or breakdown. 42 (4), pp. 447 – 465, April 2012.
[16] Krutz, R.L. & Vines R.D. Cloud Security A Comprehensive Guide to
A. Recommendation Secure Cloud Computing, U.S.A.: Wiley Publishing, Inc., 2010.
This research is a new attempt towards the development of a [17] Legris, P., Ingham, J. & Corellette, P. “Why do people use information
technology? A critical review of the technology acceptance model”,
tool and approach for studying utilization capacity and cost of

83

Authorized licensed use limited to: LIVERPOOL JOHN MOORES UNIVERSITY. Downloaded on February 07,2024 at 00:57:25 UTC from IEEE Xplore. Restrictions apply.
 Journal of Information & Management, Vol.40, Issue 3, pp. 191 – 204, [39] Zhang, S., Zhang, S., Chen, X. & Huo, X. “Cloud Computing Research
Jan. 2003 and Development Trend”, IEEE Second International Conference on
[18] Liu, W. “Research on cloud computing security problem and strategy”, Future Networks (ICFN), 2010, pp. 93 – 97.
IEEE International Conference on Consumer Electronics,
Communications and Networks (CECNet), 2012, pp. 1216 – 1219
[19] LondonPCGeek (2012) ‘Service prices tariffs and charges’ [Online].
Available: http://pcrepairs24.com/service-prices-tariffs-and-charges1
[20] Mach, W. & Schikuta, E. “A Consumer-Provider Cloud Cost Model
Considering Variable Cost”, IEEE Ninth International Conference on
Dependable, Autonomic and Secure Computing (DASC), 2011, pp.
628 – 635.
[21] Mell, P. & Grance, T. (2011) ‘The NIST Definition of Cloud
Computing: Recommendations of the National Institute of Standards
and Technology’ U.S. [Online]. Available:
http://csrc.nist.gov/publications/nistpubs/800-145/SP800-145.pdf
[22] Microsoft (2010) Security in Cloud Computing Overview [Online]
Available:
http://www.microsoft.com/download/en/details.aspx?displaylang=en&i
d=327
[23] Mikko V., Marko S. & Petri S. “Detailed cost modelling: a case study
in warehouse logistics”, International Journal of Physical Distribution
& Logistics Management, Vol. 37 (3), pp. 184 – 200, April 2007.
[24] Miller, M. Cloud Computing: Web-Based Applications That Change
the Way You Work and Collaborate Online, USA: Que Publishing,
2009.
[25] Miller, R. (2010) 'Cloud Computing as an Energy Saving Tool', Data
Centre Knowledge [Online]. Available:
http://www.datacenterknowledge.com/archives/2010/05/17/cloud-
computing-as-an-energy-saving-tool
[26] Nucleus Research (2010) “Cloud Computing: it is east to go green”
[Online]. Available: http://nucleusresearch.com/research/notes-and-
reports/cloud-computing-it-is-easy-being-green/
[27] Officegenie (2012) “Cost to rent office space” [Online]. Available at:
http://www.officegenie.co.uk/rent [Accessed February 21, 2014]
[28] Owusu, F. & Pattinson, C. “The Current State of Understanding of the
Energy Efficiency of Cloud Computing”, IEEE 11th International
Conference on Trust, Security and Privacy in Computing and
Communications (TrustCom), 2012, pp. 1948 – 1953.
[29] Patirdar, S., Rane, D. & Jain, P. “A Survey Paper on Cloud
Computing”, IEEE 2nd International Conference on Advanced
Computing & Communication Technologies, 2012, pp. 394 – 398
[30] Petri, G. (2010) “Shedding Light on Cloud Computing” [Online].
Available:
http://www.ca.com/us/~/media/files/whitepapers/mpe_cloud_primer_0
110_226890.pdf
[31] Qaisar, E.J “Introduction to cloud computing for developers: Key
concepts, the players and their offerings”, Proceedings of IEEE
Information Technology Professional Conference (TCF Pro IT) New
Jessy, 2012, pp 1-6.
[32] Ramgovind, S., Eloff, M. M. & Smith, E. “The Management of
Security in Cloud Computing”, Proceedings of IEEE Information
Security for South Africa (ISSA), 2010, pp. 1 – 7.
[33] SMART 2020: Enabling the low carbon economy in the information
age, [Online]. Available:
http://www.smart2020.org/_assets/files/02_Smart2020Report.pdf
[34] Utility Warehouse (2012) “View tariffs” [Online]. Available:
https://www.utilitywarehouse.co.uk/energy/electricity
[35] Walker, J. Accounting in a Nutshell: Accounting for the Non-specialist,
3rd Ed., U.S.A: Elsevier Ltd, 2009.
[36] Wei, Y. & Blake, M.B. “Service-Oriented Computing and Cloud
Computing: Challenges and Opportunities”, IEEE Journal of Internet
Computing, 14 (6), pp. 72 – 75, Nov. – Dec. 2010.
[37] Wickremasinghe, B., Calheiros, R.N. & Buyya, R. “CloudAnalyst: A
CloudSim-Based Visual Modeller for Analysing Cloud Computing
Environments and Applications”, IEEE 24th International Conference
on Advanced Information Networking and Applications (AINA), 2010,
pp. 446 – 452.
[38] Wu, Z. & Gan, A. “Qualitative and Quantitative Analysis the Value of
Cloud Computing”, Proceedings of the International Conference on
Information Management, Innovation Management and Industrial
Engineering (ICIII), 2011, Vol. 2, pp. 518 – 521.

84

Authorized licensed use limited to: LIVERPOOL JOHN MOORES UNIVERSITY. Downloaded on February 07,2024 at 00:57:25 UTC from IEEE Xplore. Restrictions apply.