Sirisha Potluri, Katta Subba Rao, Sachi Nandan Mohanty (Eds.

)
Cloud Security
De Gruyter Series on
Smart Computing Applications

Edited by Prasenjit Chatterjee, Dilbagh Panchal,

Dragan Pamucar, Sarfaraz Hashemkhani Zolfani

Volume 1
Cloud Security

Techniques and Applications

Edited by Sirisha Potluri, Katta Subba Rao,

Sachi Nandan Mohanty
Editors

Asst.Prof. Sirisha Potluri

ICFAI Foundation for Higher Education
Department of CSE, IcfaiTech
Faculty of Science and Technology
Donthanpally Shankarpalli Road
Telangana-501203
India
[email protected]

Dr. Katta Subba Rao

B.V.Raju Institute of Technology
Department of CSE
Medak District
Telangana-502313
India
[email protected]

Dr. Sachi Nandan Mohanty

College of Engineering Pune
Department of Computer Engineering
Wellesley Road, Shivajinagar, Pune
Maharashtra-411005
India
[email protected]

ISBN 978-3-11-073750-9
e-ISBN (PDF) 978-3-11-073257-3
e-ISBN (EPUB) 978-3-11-073270-2
ISSN 2700-6239

Library of Congress Control Number: 2021939885

Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed
bibliographic data are available on the Internet at http://dnb.dnb.de.

© 2021 Walter de Gruyter GmbH, Berlin/Boston

Cover image: monsitj / iStock / Getty Images Plus
Printing and binding: CPI books GmbH, Leck

www.degruyter.com
|
This book is humbly dedicated in honour of Shri N. J. Yasaswy, Finance and invest-
ment writer and founder of ICFAI University and Shri E. N. Murthy, Member Secre-
tary, The ICFAI Society, whose support, encouragement and blessings made us to
get this effort and work.

Aforesaid volume contains the devotion with great love and affection “In memory of
my beloved father Shri Potluri. Rama Mohana Rao”, whose foundation, inspiration
and benediction made me to achieve this great career and life.
Preface
Sustainable computing paradigms like cloud and fog are mining better solutions to
handle the issues of security, privacy, performance, storage, processing, mainte-
nance, efficiency, integration, cost, energy and latency. In these evolving days,
according to statistics, billions of connected devices are producing enormous
amounts of real time data. For complex computation and processing of the collected
data in order to make the dynamic decisions these devices are connected over the
cloud or fog environment. Best practices and protocols are being implemented by
cloud to ensure security and privacy everywhere in the architecture.
This book illuminates some of the best practices and their respective outcomes
in cloud environment to preserve privacy and security. This book is aiming to pro-
vide next level of security and privacy preserving techniques for dynamic cloud.
Innovative techniques and methods for secure cloud are gathered to present ad-
vanced and specialized research in the selected area.
Various state-of-the-art and cutting edge methods to achieve security and priva-
cy in cloud computing environment by using Artificial intelligence, Machine learn-
ing, Blockchain, IoT, DDoS mitigating practices, Industry 4.0, Cloud manufacturing,
Data analytics and Vanet are essential slices of this book.
This book essences on various research issues related to security and privacy
preserving approaches and techniques using big data storage and analysis, large
scale data processing, knowledge discovery and knowledge management, computa-
tional intelligence, data security and privacy, data representation and visualization
and data analytics. The featured technology of this book optimizes various industry
processes using business intelligence in engineering and technology. This book
sheds light on cloud based secure integrated and development practices to increase
productivity and reduce the operational cost.
Chapter 1 focuses on AI based advanced technology that helps companies to
manage complex operations efficiently and improve productivity. AI-based net-
works allows to maximize organizational performance but also to evade risks and
threats. This chapter focuses on security and privacy related issues and explored
effective counter measure to handle such issues. This chapter also provides latest
literature on collaboratively secured collections to identify the kind of vulnerability
they address and also the form of approach they use to handle it.
Chapter 2 focuses on Blockchain based advanced technology to deliver infra-
structure and has been adopted widely due to its performance and availability.
Blockchain Technology, one of the emerging technologies can be used to comple-
ment the issues concerning cloud users. This technology provides features like de-
centralization, distribution, authenticity, immutability, trust, etc. that can be used
to face the threats of data integrity. This chapter focuses on concerning the security

https://doi.org/10.1515/9783110732573-202
VIII | Preface

and privacy of the data in the cloud model and how the features of the blockchain
technology can enhance the integrity of this model.
Chapter 3 focuses on DDoS attack occurrences, DDoS types and associated secu-
rity and privacy issues in cloud computing environment. Due to the dynamic behav-
ior, distributed paradigm and heterogeneity present among the processing ele-
ments, devices and service oriented pay per use policies; the cloud computing
environment is having its availability, security and privacy issues. Among these
various issues one of the important issues in cloud computing paradigm is DDoS
attack. This chapter put in plain words the DDoS attack, its detection as well as
prevention mechanisms in cloud computing environment. This chapter also ex-
plains about the effects of DDoS attack on cloud platform and the related defence
mechanisms required to be considered.
Chapter 4 focuses on secure cloud based healthcare application to handle big
data of healthcare industry. Handling and processing such huge chunks of data can
be a daunting task. To avoid scenarios of this type, many of the healthcare sectors
have adapted cloud computing as the solution. This chapter explains privacy relat-
ed issues and security related concerns, dynamically acceptable storage, regulatory
issues, electronic documents containing patient’s data, handwritten medical edi-
tions, various images such as X-ray, MRI scan and radiology images analysis.
Chapter 5 focuses on secure IoT frameworks, methods, security arrangements,
and the best protection models are important and suitable for the various layers of
IoT operated applications. This chapter also proposed embossed IoT model: stand-
ard and expanded with security and safety components and layers of physical evi-
dence. This chapter also explains about perceived security vulnerabilities and best
security measures to combat network security risks in all layers of cloud, edge, and
IoT.
Chapter 6 focuses on efficient marketing strategies for business to business
marketing and business to customer marketing are greatly heightened by using
cloud based marketing automation solutions. Cloud governance ensure to meet
business desires and needs through professional practices. Significant factors to
ensure security in cloud marketing are embedded in each phase of marketing life
cycle. This chapter explains about various cloud based secure marketing solutions
to meet next generation marketing needs.
Chapter 7 focuses on secure machine learning model to anticipate in finest out-
comes and boost the choice of cloud based long-term course of actions. This chapter
is proposing best strategic cloud security enhancement model for next generation
computing standards. Efficient machine learning algorithms like convolution neural
network gives automatic and responsive approaches to reinforce security in a cloud
environment. These models give solutions that incorporate holistic approaches for
secure enterprise knowledge throughout all the cloud applications.
Chapter 8 focuses on secure intelligent VANET for smart transportation to ob-
serve resolutions using cloud based infrastructure. The vehicles connect together
 Preface | IX

through a network to form a vehicular cloud which will offer space for storing, com-
puting resources, sensor readings as an on-demand service to clients. There is a
need of efficient computational, supervising and guiding solutions for the citizens
to make them to get rid of traffic related issues and problems. This chapter explains
various methods to find an appropriate parking lot within the area on the brink of
university, shopping complex, or a commercial complex in big and medium cities
would enjoy the assistance of an automatic parking management utility.
Chapter 9 focuses on novel cloud manufacturing architype which is developed
from prevailing innovative manufacturing prototypes and enterprise level infor-
mation expertise under the provision of cloud computing, IoT, virtualization tech-
nology and service oriented computing, and cutting-edge computing technologies
and expertise. Manufacturing as a service is an advanced manufacturing architype
developed from the existing manufacturing models such as ASP, AM, NM and
MGrid. This chapter is to study and discuss about security and privacy issues and
solutions in cloud based manufacturing.

Sirisha Potluri, India

Katta Subba Rao, India
Sachi Nandan Mohanty, India
Acknowledgement
The editors would like to acknowledge and pass on good wishes and show apprecia-
tion to all the authors for contributing their chapters. We would also like to thank
the subject matter experts who could find their time to review the chapters and de-
liver those in time. Our special thanks to the people who gave their time to advice
and suggest in refining our thoughts and approaches accordingly to produce richer
contributions. We are particularly grateful to the Walter de Gruyter GmbH publish-
ers for their amazing crew to support us in all ways of encouragement, engagement,
support, cooperation and contribution to publish this book.

https://doi.org/10.1515/9783110732573-203
Contents
Dedication | V

Preface | VII

Acknoledgment | XI

List of Abbreviations | XV

List of Contributors | XVII

Sachi Nandan Mohanty, Sirisha Potluri, V. Bhanu Prakash Reddy, B.Srinath, B.

Manjunath Reddy
Cloud Security Concepts, Threats and Solutions: Artificial Intelligence Based
Approach | 1

Shagun S Lokre, Shanmukhi Priya, Vihas Naman, Sachi Nandan Mohanty, Sirisha
Potluri
Addressing Security and Privacy in Cloud Computing: Blockchain as a
Service | 21

Sirisha Potluri, Sachi Nandan Mohanty, T. Kundana, D. Abhinav, P. Sushrutha

Security and Privacy Preservation Model to Mitigate DDoS Attacks in Cloud | 41

Sirisha Potluri, Sai Lalitha Sunaina, Neha Pavuluri, Chennu Sai Sri Govind,
Raghavender Rao Jakileti, V. MNSSVKR Gupta
A Secure Cloud Infrastructure towards Smart Healthcare: IoT Based Health
Monitoring | 63

Indrani Inapakolla, SVB Revanth, Siliveru Akhil Durga, Sirisha Potluri, Sachi Nandan
Mohanty
Internet of Cloud: Secure and Privacy Preserving Cloud Model with IoT Enabled
Service | 83

Srikanth Pothuri
Marketing analytics as a Service: Secure Cloud Based Automation Strategy | 105

Sachi Nandan Mohanty, Gouse Baig Mohammad, Sirisha Potluri, Ramya Reddy
Padala, Lavanya Reddy Padala
Next Generation Cloud Security: State of the Art Machine Learning Model | 125
XIV | Contents

Sirisha Potluri, Gouse Baig Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K.

Sahaja
Secure Intelligent Framework for VANET: Cloud Based Transportation
Model | 145

Sirisha Potluri, Sachi Nandan Mohanty, A. D. Sriram Kumar, D. Maheswari, B.

Rahini
Cloud Manufacturing Service: A Secure and Protected Communication
System | 171

Index | 191
List of Abbreviations
AI Artificial Intelligence
DI Data Integrity
DU Data Usage
TTP Third Party Based Encryption Scheme
VM Virtual Machine
CDC Centralized Data Centre
ANN Artificial Neural Network
MSE Mean Squared Error
DNN Deep Neural Network
IoT Internet of Things
CSP Cloud Service Provider
SAMS Secure Authentication Management human-centric Scheme
DDoS Distributed-Denial-of-Service
DoS Denial-of-service
SLA Service Level Agreement
SDN Software Defined Network
KNN K nearest Neighbour
SOA Service Oriented Architecture
BW-DDoS Bandwidth distributed Dos
CDN Content Delivery Network
SMC Secure Multiparty Computation
API Application Programming Interface
VMM Virtual Machine Manager
SaaS Software as a Service
PaaS Platform as a Service
IaaS Infrastructure as a Service
BDaaS Big Data as a Service
NaaS Network as a Service
HaaS Healthcare as a Service
MaaS Manufacturing as a Service
IoTaaS IoT as a Service
AaaS Analytics as a Service
EMS Emergency Medical Systems
CNN Convolutional Neural Network
AWS Amazon Web Service
RFID Radio Frequency Identification
IAM Identity and Access Management
SEO Search Engine Optimization
SEM Search Engine Marketing

https://doi.org/10.1515/9783110732573-205
XVI | List of Abbreviations

CRM Customer Relationship Management

CapEx Capital Expense
OpEx Operating Expenditure
RACE Reach, Act, Convert, and Engage
CRO Conversion Rate Optimization
LPO Landing Page Optimization
SMB Small and Medium Sized Businesses
BFSI Financial Services and Insurance
PPC Pay-Per-Click
CTA Call to Action
ML Machine Learning
CAGR Compound Annual Growth Rate
SIEM Security Information and Event Management System
VCC Vehicular Cloud Computing
VANET Vehicular Ad hoc Network
LTE Long Term Evolution
WAVE Wireless Access Vehicular Environment
DSRC Dedicated Short Range Communication
BSM Basic Safety Message
QoS Quality of Service
LIDAR Laser Illuminated Detection and Ranging
SARTRE Safe Road Trains for the Environment
VC Vehicular Cloud
PKI Public Key Infrastructure
CRL Certificate Revocation List
CA Certificate Authority
ITS Intelligent Transportation Systems
ECPP Enhanced Conditional Privacy Preservation
PACP Pseudonymous Authentication-Based Conditional Privacy
List of Contributors
Ms. Sirisha Potluri working as Assistant Professor in the Department of Computer Science & Engi-
neering at ICFAI Foundation for Higher Education, Hyderabad. She is pursuing her PhD in the area of
cloud computing from KL Education Foundation, Vijayawada, India. Her research areas include
Distributed Computing, Cloud Computing, Fog Computing, Recommender Systems and IoT. She has
7+ years of teaching experience. Her teaching subjects include Computer Programming using C,
Data Structures, Core JAVA, Advanced JAVA, OOP through C++, Distributed Operating System, Hu-
man Computer Interaction, C# and .NET Programming, Computer Graphics, Web Technology, UNIX
Programming, Distributed and Cloud Computing, Python Programming and Software Engineering.
She has published 22 scholarly peer reviewed research articles in reputed International Journals.
She has attended 11 international conferences organized by various professional bodies like IEEE
and Spinger to present her novel research papers. She has a patent in cloud computing domain and
edited two books. She is a member of IEEE Computer Society. Sirisha is a reviewer for various jour-
nals of Inderscience Publishers and reviewed many manuscripts.

Dr. Sachi Nandan Mohanty , Associate Professor in the Department of Computer Engineering at the
College of Engineering Pune, India. He received his PostDoc from IIT Kanpur in the year 2019 and
Ph.D. from IIT Kharagpur, India in the year 2015, with MHRD scholarship from Govt of India. He has
edited 14 books in association with Springer, Wiley and CRC Press. His research areas include Data
mining, Big Data Analysis, Cognitive Science, Fuzzy Decision Making, Brain-Computer Interface, and
Computational Intelligence. Prof. S N Mohanty has received 3 Best Paper awards during his Ph.D at
IIT Kharagpur from International Conference at Benjing, China, and the other at International Con-
ference on Soft Computing applications organized by IIT Rookee in the year 2013. He has awarded
Best thesis award first prize by Computer Society of India in the year 2015. He has published 42
International Journals of International repute and has been elected as FELLOW of Institute of Engi-
neers, IETE, and senior member of IEEE Computer Society Hyderabad chapter. He also the reviewer
of Journal of Robotics and Autonomous Systems (Elsevier), Computational and Structural Biotech-
nology Journal (Elsevier), Artificial Intelligence Review (Springer), Spatial Information Research
(Springer).

Dr. Katta Subba Rao is working as a Professor in Department of Computer Science & Engineering in
B.V.Raju Institute of Technology, Hyderabad, India. He has received PhD degree from Nagarjuna
University, Andhra Pradesh, India. His research areas includes Software Engineering, Cloud Compu-
ting and Machine Learning. He has more than 15 years of cherished teaching and research experi-
ence. His research contributes various journal articles, conferences and book chapters.

Dr. Gouse Baig Mohammad is working as Associate Professor in Department of Computer Science &
Engineering in Vardhaman College of Engineering, Hyderabad, India. He has received PhD degree
from Nagarjuna University, Andhra Pradesh, India. His research areas includes Network Security,
IoT and Machine Learning. He has 14 years of cherished teaching and research experience. His
research contributes 7 scholarly peer reviewed research articles in reputed International Journals
and 2 patents.

Mr. Srikanth Pothuri is working as National Head, Department of Marketing, at ICFAI Foundation for
Higher Education, Hyderabad, India. He is pursuing his PhD from ICFAI Business School, ICFAI Uni-
versity Dehradun, Dehradun, Uttarakand, India. His research areas include Marketing Management,
XVIII | List of Contributors

Marketing Strategy, Consumer Behavior and Marketing Quantitative Methods. He has 19 years of
industry experience.

Mr. V. MNSSVKR GUPTA is working as Assistant professor in Sagi Rama Krishnam Raju Engineering
College (SRKR Engineering College), Bhimavaram, India. He is pursuing his PhD from KL Education
Foundation, Vijayawada, India. His research areas include Artificial Intelligence, Bio Informatics and
Image Processing. He has 12 years of teaching experience. He has published 11 scholarly peer
reviewed research articles in reputed International Journals and attended 5 international confer-
ences.

Mr. Vangeti Bhanu Prakash Reddy is a researcher in the Department of Computer Science and
Engineering at ICFAI Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology),
Hyderabad, India. His research interest includes Artificial Intelligence, Cloud Computing and Ma-
chine Learning.

Mr. Srinath Bellamkonda is a researcher in the Department of Computer Science and Engineering at
ICFAI Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad,
India. His research interest includes Artificial Intelligence, Cloud Computing and Machine Learning.

Mr. Manjunath Reddy is a researcher in the Department of Computer Science and Engineering at
ICFAI Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad,
India. His research interest includes Artificial Intelligence, Cloud Computing and Machine Learning.

Mr. Shagun S Lokre is a researcher in the Department of Computer Science and Engineering at ICFAI
Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad, India.
His current area of interest for research includes Blockchain technology, Ethereum, Distributed
Ledger Technology, Supply chain management. As a part of his achievements, he has an approved
patent in the field of Blockchain technology in the event management sector in the India patent
office (IPO), and co-authored 3 chapters titled “Gun Tracking System Using Blockchain Technology”,
“Secure Event Ticket Booking Using Blockchain Technology”, “Distributed Ledger Technology in the
Construction Industry Using Corda” in Blockchain Technology. These chapters are part of “Block-
chain Technology: Applications and Challenges”, “Intelligent Systems Reference Library, Springer
(WoS and Scopus Indexed)”, and “The New Advanced Society: Artificial Intelligence and Industrial
Internet of Things Paradigm, Willey Press (U.S.A)”. He is currently working on other research papers
on the Distributed Ledger Technology.

Ms. Shanmukhi Priya, is a researcher in the Department of Computer Science and Engineering at
ICFAI Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad,
India. Her current area of interest for research includes blockchain technology, Ethereum, Distribut-
ed Ledger technology, Corda, and back-end technologies. As a part of his achievements, he has an
approved patent in the field of Blockchain technology in the event management sector in the India
patent office (IPO), and co-authored 3 chapters titled “Gun Tracking System Using Blockchain Tech-
nology”, “Secure Event Ticket Booking Using Block-chain Technology”, “Distributed Ledger Tech-
nology in the Construction Industry Using Corda” in Blockchain Technology. These chapters are part
of “Blockchain Technology: Applications and Challenges”, “Intelligent Systems Reference Library,
Springer (WoS and Scopus Indexed)”, and “The New Advanced Society: Artificial Intelligence and
Industrial Internet of Things Paradigm, Willey Press (U.S.A)”. He is currently working on other re-
search papers on the Distributed Ledger Technology.
 List of Contributors | XIX

Mr. Vihas Naman is a researcher in the Department of Computer Science and Engineering at ICFAI
Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad, India.
His current area of interest for research includes blockchain technology, Ethereum, Distributed
Ledger technology, Corda, and the Internet of things. As a part of his achievements, he has an
approved patent in the field of Blockchain technology in the event management sector in the India
patent office (IPO), and co-authored 3 chapters titled “Gun Tracking System Using Blockchain Tech-
nology”, “Secure Event Ticket Booking Using Block-chain Technology”, “Distributed Ledger Tech-
nology in the Construction Industry Using Corda” in Blockchain Technology. These chapters are part
of “Blockchain Technology: Applications and Challenges”, “Intelligent Systems Reference Library,
Springer (WoS and Scopus Indexed)”, and “The New Advanced Society: Artificial Intelligence and
Industrial Internet of Things Paradigm, Willey Press (U.S.A)”. He is currently working on other re-
search papers on the Distributed Ledger Technology.

Ms. T. Kundana is a researcher in the Department of Computer Science and Engineering at ICFAI
Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad, India.
Her research interest includes Network Security, Cloud Computing, Machine Learning and Cyber
Security.

Mr. D. Abhinav is a researcher in the Department of Computer Science and Engineering at ICFAI
Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad, India.
His research interest includes Network Security, Cloud Computing, Machine Learning and Cyber
Security.

Ms. P. Sushrutha is a researcher in the Department of Computer Science and Engineering at ICFAI
Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad, India.
Her research interest includes Network Security, Cloud Computing, Machine Learning and Cyber
Security.

Ms. Sai Lalitha Sunaina M is a researcher in the Department of Computer Science and Engineering
at ICFAI Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad,
India. Her research interest includes Cloud Computing, Machine Learning and Data Analytics.

Ms. Neha Pavuluri is a researcher in the Department of Computer Science and Engineering at ICFAI
Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad, India.
Her research interest includes Cloud Computing, Machine Learning and Data Analytics.

Mr. Govind Ch is a researcher in the Department of Computer Science and Engineering at ICFAI
Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad, India.
His research interest includes Cloud Computing, Machine Learning and Data Analytics.

Mr. Raghavender Rao Jakileti is a researcher in the Department of Computer Science and Engineer-
ing at ICFAI Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hydera-
bad, India. His research interest includes Cloud Computing, Machine Learning and Data Analytics.

Ms. Indrani Inapakolla is a researcher in the Department of Computer Science and Engineering at
ICFAI Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad,
India. Her research interest includes Cloud Computing, Data Analytics and Data Visualization.
XX | List of Contributors

Mr. SVB Revanth is a researcher in the Department of Computer Science and Engineering at ICFAI
Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad, India.
His research interest includes Cloud Computing, Data Analytics and Data Visualization.

Mr. Siliveru Akhil Durga is a researcher in the Department of Computer Science and Engineering at
ICFAI Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad,
India. His research interest includes Cloud Computing, Data Analytics and Data Visualization.

Mr. A. D. Sriram Kumar is a researcher in the Department of Computer Science and Engineering at
ICFAI Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad,
India. His research interest includes Cloud Computing, Machine Learning, Industry 4.0 and IoT.

Ms. D. MAHESHWARI is a researcher in the Department of Computer Science and Engineering at

ICFAI Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad,
India. Her research interest includes Cloud Computing, Machine Learning, Industry 4.0 and IoT.

Ms. B. Rahini is a researcher in the Department of Computer Science and Engineering at ICFAI Foun-
dation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad, India. Her
research interest includes Cloud Computing, Machine Learning, Industry 4.0 and IoT.

Ms. Ramya Reddy P is a researcher in the Department of Computer Science and Engineering at ICFAI
Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad, India.
Her research interest includes Cloud Computing, Machine Learning and Artificial Intelligence.

Ms. Lavanya Reddy P is a researcher in the Department of Computer Science and Engineering at
ICFAI Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad,
India. Her research interest includes Cloud Computing, Machine Learning and Artificial Intelligence.

Ms. M. Vaishnavi is a researcher in the Department of Computer Science and Engineering at ICFAI
Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad, India.
Her research interest includes Cloud Computing, Machine Learning, Vanet and Network Security

Ms. K. Sahaja is a researcher in the Department of Computer Science and Engineering at ICFAI
Foundation for Higher Education, IcfaiTech (Faculty of Science and Technology), Hyderabad, India.
Her research interest includes Cloud Computing, Machine Learning, Vanet and Network Security
Sachi Nandan Mohanty, Sirisha Potluri, V. Bhanu Prakash, B.
Srinath, B. Manjunath
Cloud Security Concepts, Threats and
Solutions: Artificial Intelligence Based
Approach
Abstract: Artificial intelligence is an advanced technology that helps companies to
manage complex operations efficiently and improve productivity. In this genera-
tion, many business communities are using AI-based networks to maximize organi-
zational performance but also to evade risks and threats. Security and privacy are
both major issues with AI technology that contributes to data privacy and leads to
hacking problems. The proposed study focused on security issues of AI technology
and explored effective counter measures to improve data privacy. Privacy-
preserving cooperative filtering recommendation systems are aiming to supply users
with correct recommendations to maintain bound assurances regarding the privacy
of their information. This survey examines the latest literature on collaboratively
secured collections, provides an overview of the sector and distinguishes significant
contributions. The study also focuses and comforts to identify the kind of vulnera-
bility they address and also the form of approach they use to handle it.

Keywords: Cloud Computing, Cloud Security, User Identification System, Data

Storage, Data Integrity, Artificial Intelligence

||
Sachi Nandan Mohanty, Department of Computer Engineering, College of Engineering Pune,
Wellesley Rd, Shivajinagar, Pune, Maharashtra 411005, India, [email protected]
Sirisha Potluri, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana
501203, India, [email protected]
V. Bhanu Prakash Reddy, Department of CSE, IcfaiTech (Faculty of Science and Technology),
ICFAI Foundation for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad,
Telangana 501203, India, [email protected]
B. Srinath, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana
501203, India, [email protected]
B. Manjunath Reddy, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana
501203, India, [email protected]

https://doi.org/10.1515/9783110732573-001
2 | Sachi Nandan Mohanty, Sirisha Potluri, V. Bhanu Prakash, B. Srinath, B. Manjunath

1 Introduction
Cloud Computing is one of the pioneers to add to the field of computer technology in
recent times to redesign the possession value using pay-as-you-go model. However,
in terms of security, the current methods: anti-virus programs, authentication
mechanisms, firewall safety features do not appear to be able to withstand the se-
verities of threats.
Therefore, the user identification system, registers user activity to investigate
system performance, adds a protection system in an efficient and collaborative way
which is combined with latest practices to provide increased security. Our work
focuses on developing an efficient user identification program to exploit the envi-
ronment in normal way and course of conduct (of the user identification program)
and proposes an efficient replacement hybrid approach, which will bring a compre-
hensive user safety identification program in cloud computing [1-3].

Artificial
Internet of things
intelligence

Cloud computing

Blockchain

Fig. 1: Emerging research areas of cloud computing

The utilization of proposal frameworks has improved significantly in recent times.

Shopping malls present their clients recommendations with suggestions based on
 Cloud Security Concepts, Threats and Solutions: Artificial Intelligence Based Approach | 3

set of past experiences and socio-economic status; Film and book recommendation
sites are consistently used to select new top choices of their viewers and readers;
online music streaming management generates dynamic playlists to suit the trends
of their listeners.
It is magnificent that the new era of history and new technological change is on
the rise in every platform. We tend to believe that a new era of “Internet and artifi-
cial intelligence (AI)”, characterized by ubiquitous networks, data technologies,
shared services, border integration, automation, and mass production is coming.
The rapid development and integration of the latest AI and web technologies, new
data technologies, new energy technologies, mechanical technologies, and biotech-
nology are a necessary requirement in this new era, which can transform models,
methods and natural resources, welfare and national security.
Cloud platform provides significant services such as software package resources
with many benefits such as data storage, server maintenance and accessibility for
cloud tip users using data access. Due to the rising popularity of the cloud, a large
number of databases have been exported to reduce the amount of storage and cost.
In addition, data integrity (DI) is an ongoing problem in the cloud to deal with data
security and data consistency. Therefore, data usage (DU) face the risk of using
cloud computing for storage and data transfer applications, due to the shortage of
DI. For the purpose of partitioning the DI problem, the cloud has a high reliability
band using trusted third party based encryption scheme (TTP). The TTP function is
to ensure continuous DU authentication with continuous DI protection. To ensure
data privacy, cloud must use efficient data encryption methods.
Cloud Computing is an expanding field in the area of computing. It is a way to
increase resource usage and computing power without having the financial burden
of buying new infrastructure on the users. Cloud computing services witnesses
many features such as efficiency, reliability, flexibility, profitability, elasticity,
availability and on demand assets delivery using utility computing. Cloud resources
are usually measured based on their usage to complete the execution of the tasks in
the given assignment.
Cloud computing provides services through internet and web applications. Re-
sources dynamically go up or down based on the demand and in addition to that
cost efficient pricing is followed by a simple payment policy by using pay as you go
model. Therefore, cloud computing platform has several resources which are pro-
vided by cloud service provider and the cloud storage processes are changing dy-
namically to meet the demands of ongoing data and information growth. However,
the benefits of cloud storage go progressively with a group of cyber security related
protocols and methods.
The privacy issue is one of the most serious threats to data loss, malicious alter-
ations, server crashes with other cyber threats samples such as Yahoo’s 3 billion
account exposure by hackers in 2013, Apple’s iCloud leak in 2014, infringement of
information privacy of Drop box In 2016, a leaked iCloud event, wherever the vari-
4 | Sachi Nandan Mohanty, Sirisha Potluri, V. Bhanu Prakash, B. Srinath, B. Manjunath

ous photos of Hollywood actors were unveiled and caused a lot of outcry. Those
events have a profound effect on the company’s reputation.
Fog integration includes the cloud computing paradigm to the edge of the net-
work, thus providing a new type of applications, services and security in the cloud
environment. Therefore, we are often ready to develop cloud technology with pro-
tected and secured user information using artificial intelligence mechanisms.

2 Literature Analysis
Artificial intelligence is one of the siginificant research area related to cloud
computing. Emerging research areas of cloud computing are artificial intelligence,
blockchain and internet of things as shown in fig. 1. The importance and
significance of our research is to determine safety risks and problems associated
with AI technology and explore the results of recent studies.

2.1 Significance of artificial intelligence

Čerka, Grigienė, and Sirbikytė, (2015) examined that security is a necessary part of
AI technology to detect and handle whenever hackers transmit unwanted signals
and reduces the data privacy and security related issues. Within organizations, AI
technology helps to perform tasks effectively but the presence of unauthorized
signals will violate data privacy and lead to data breach problems.
Three key expertise that lead to privacy preservation using artificial intelligence
are deep learning, natural language processing, machine learning, neural networks
and computer vision. Improper configuration of unauthorized activities and services
owned by various cloud devices and infrastructure are efficiently addressed by AI
based mechanisms. It is important to use secure networks to handle malware
transmission in cloud computing environement where the privacy of AI networks
can be increased and the performance of advanced systems can be enhanced.
Guihot, Matthew, and Suzor, (2017) provided their views and stated that AI
technology helps to improve business process, and to handle data related risks,
challenges and breaches in organizations. Network based AI uses the internet
connection and communication systems where some times the presence of
unauthorized signals can lead to cybercrime and results into reducing the privacy of
sensitive information. If the malware is transmitted and enters into AI based cloud
systems, identification, control and handling such transimission is handled by
using AI based mechanisms [4-7].
 Cloud Security Concepts, Threats and Solutions: Artificial Intelligence Based Approach | 5

2.2 Security issues

Pan, (2016) determined that there are variety of security risks or issues that occur
within the AI networks like security threat of technology abuse, induced by tech-
nical defects and plenty of similar issues. It’s necessary to seek out the danger fac-
tors coupled with cloud computing environment so as to defend these cyber-crimes
and risks from the developed systems using AI technology [8].

2.3 Privacy issues

According to Rehman, and Saba, (2014) the privacy of artificial intelligence is often
broken by interference communication channels and accessing personal networks
of the organizations. Within the advent of artificial intelligence, various privacy
risks such as confidentiality problems, phishing, and unauthorized signals trans-
ferred by the criminals are handled more efficiently when compared with the exist-
ing solutions. Therefore, it is significant to specialize in privacy whereas implement-
ing AI networks in the organizations should guarantee data security and prefer to
use solely private networks instead of third parties [9].

2.4 Security and privacy enhancement using artificial

intelligence
Artificial Intelligence (AI) goal is to create IoT based infrastructure and fog nodes
that monitor the work environment and continuously adapt to provide superior QoS
features, reduce the consumption of power and total infrastructure cost.
AI constitutes of various search algorithms, machine learning, and reinforce-
ment learning to efficiently handle issues and problems in cloud. In today’s world of
deep data operations with fog growth and cloud deployment, more and more AI is
needed at different levels to provide better workflow decisions, VM migration, etc. to
improve the previously mentioned conditions.
There are many projects aimed at investing in AI strategies to increase the per-
formance of cloud and fog systems. Completely different functions apply directly to
cloud planning policies, virtualization algorithms, distribution systems etc. which
uses the search methods such as genetic algorithms, supervised machine learning
and robust in-depth learning to perform their intended functions.
AI delivers an effective way to optimize large systems with more detail in sim-
plicity of engineering and efficiency by permitting automated decision-making ra-
ther than man-made solutions that provide more efficient and faster resolutions.
With the current expansion of the internet, the existence of sensors in the uni-
verse, the emergence of big data, the development of e-commerce, the growth of the
6 | Sachi Nandan Mohanty, Sirisha Potluri, V. Bhanu Prakash, B. Srinath, B. Manjunath

information society, and the integration of data and information with the society,
the physical environment, and cyberspace, the information environment: AI 2.0
(Pan, 2016) observes great changes in cloud. The arrival of latest technologies em-
powers a new phase of AI. Key features of AI 2.0 include the emergence of visual
acuity driven by in-depth learning data, internet-based intelligence, human and
technology-based mechanisms, and high-level media thinking (Pan, 2016).
The continued emergence of smart cities, robots in medical care, automated
transport, virtual assistants, mechanical device, automated cars, smart phones,
smart toys, smart communities and smart economies provide a wider market de-
mand which urges for the new developments in both AI technologies and applica-
tions.
The functions of learning, understanding, making decisions, solving problems
and spontaneous thinking such as human thinking is called Artificial Intelligence.
The various methods of AI are: fuzzy programs (powerful for language representa-
tion), artificial neural networks (which mimic the structure of the human brain;
contain highly complex connections of simple mathematical units representing
complex problem-solving tasks), genetic algorithms (on purpose of the evolution of
human genes to find the right solution) etc. provides machine learning and automa-
tion solutions for problem solving [10].
Cloud computing is growing rapidly, and CDC (Centralized Data Center) have
become an integral part of prominent industries such as Amazon, Apple, Microsoft,
Google, IBM, Facebook. However, it is burdensome to monitor the operations of
large data centers manually.
Yotascale is a next-generation computing and automated performance monitor-
ing solution to reduce human response. Yotascale uses historical data to make pre-
dictions or decisions about cloud costs using artificial intelligence and helps to save
lot of expenditure. In addition, real-time analysis can be done using Yotascale to
detect unfavorable trends using in-depth reading strategies (super-
vised/unsupervised or forecasting models) and find the cause and provide future
predictions based on cloud usage and its cost. Research specialization in distributed
and cloud computing are given shown in fig. 2 and fig. 3.
In Artificial Intelligence, ANN- artificial neural network is one of the best classi-
fication method can be used as an intrusion detection system. Artificial neural net-
work is an efficient information processing method that is motivated by the natural
and biological nervous structure. It tries to signify the physical brain, functionality
and its thinking process by means of a network, electronic circuit and set of pro-
grams.
 Cloud Security Concepts, Threats and Solutions: Artificial Intelligence Based Approach | 7

Distributed Computing

Cloud computing

Fog computing Application design Data processing

Quantum computing Fault tolerance Cloud containers
Serverless computing 5G and beyond QoS and SLA

Fig. 2: Research specialization of distributed computing

Cloud Computing

Utility computing

Software engineering
Bitcoin currency Software defined network
Big data analytics and
Security and privacy Energy management
data science
Resource management Autoscaling
Deep learning

Fig. 3: Research specialization of cloud and utility computing

AI technology facilitates the development of new models, methods, and forms, sys-
tem design, and technology systems in the field of intelligent production. AI is used
in the intelligent manufacturing sector with an intelligent production system. The
use of AI beyond the intelligent production system makes no sense. Against the
8 | Sachi Nandan Mohanty, Sirisha Potluri, V. Bhanu Prakash, B. Srinath, B. Manjunath

background of ‘Internet plus AI’, an intelligent production system characterized by

independent intelligence, communication, collaboration, learning, analysis, under-
standing, decision-making, control and humanization, machine, material, environ-
ment and knowledge throughout the system and life cycle.

3 Security
Security is a major factor to consider while adapting to the cloud. Users will store a
lot of secure information on their computers. When using cloud computing technol-
ogy this data will be transferred from their computer to the cloud. Therefore, the
cloud must have effective security measures to protect this data. Threats and attacks
observed in cloud computing are shown in fig. 4.
Pan, (2016) determined that there are a number of security risks or problems
that occur in the AI networks such as security threat of technology abuse, induced
by technical defects and many more. It is important to find the risk factors linked
with AI technology in order to defend cyber-crimes and risks from the developed
systems. From recent literature, it is found that AI is a neutral technique that con-
nects numbers of technologies and computing devices to each other for performing
operations effectively [11].
The following are list of security issues observed in cloud computing and same
can be efficiently addressed by artificial intelligence practices.
– Control management
– Risk management
– Compliance
– Vulnerability management
– Patch management
– Physical security
– Data security
– Operational security
– Legal issues
– Identity management
– Application security
– Storage security
– Network security
– Information security
– Cyber security
– Virtualization security
– Auditing and maintenance
– Data encryption and key generation
 Cloud Security Concepts, Threats and Solutions: Artificial Intelligence Based Approach | 9

Cloud components
- Virtual machine
- Web server
- Operating system
- Application
- Data sources

Threats
- Spoofing
- Tampering
- Repudiation
- Information disclosure

Threats - Denial of service

and attacks - Elevation of privilege

in cloud
computing Attacks
- Abuse functionality
- Data structure attack
- Embedded malicious code
- Exploitation of authentication
- Injection
- Path traversal attack
- Probabilistic techniques
- Protocol manipulation
- Resource depletion
- Resource manipulation
- Sniffing attack
- spoofing

Fig. 4: Threats and attacks observed in cloud computing

Most of the companies suffer cyber-attacks due to technology abuse where the crim-
inals transfer unwanted signals and impact on the developed systems by which the
sensitive data can be leaked. Pesapane, et al., (2018) identified that criminals use
10 | Sachi Nandan Mohanty, Sirisha Potluri, V. Bhanu Prakash, B. Srinath, B. Manjunath

malicious programs for transferring fraud signals that help to reduce the perfor-
mance of the communication channels and technologies used in the AI systems are
not able to detect unwanted signals by which hackers can perform data breach.
Because of some technical defects, the AI networks are not capable to protect data
against cyber-crimes and more than 34% of the companies are facing such prob-
lems.
Ransbotham, et al., (2017) supported the argument and examined that lack of
awareness is a key problem where companies do not provide complete training to
the employees wherein hackers can obtain login credentials of computing devices
using malicious programs.
Cloud computing has advantages such as easy implementation, accessibility,
distribution, reliability, error tolerance, shared resources, increased storage capaci-
ty and cost-saving technology. While Cloud computing has many advantages, it still
suffers with many security issues and violations with respect to both cloud service
providers and cloud users.
According to a Cloud Security Alliance article, “Data breaches, unfortunate in-
formation, recording or administrative tasks, unauthorized integration with Appli-
cation Programming Interfaces (APIs), DoS attacks, bad content, cloud abuse, inef-
ficiency due to inefficiency, eventually shared new risks”, are identified as the top
nine threats to cloud computing. Significance of AI in cyber security is shown in
fig. 5.

Reconnaissance C & C*2 Intrusion expansion

Weaponization Installation Action on objective

Delivery Exlpoitation

Fig. 5: Significance of AI in cyber security

 Cloud Security Concepts, Threats and Solutions: Artificial Intelligence Based Approach | 11

4 Privacy
Privacy and security often confuse words. Data security ensures that data is available
when those with authorized access need it. Data privacy ensures that data is used
properly. Privacy is defined as the individual’s right to determine how and to whom
information is transmitted. Data privacy can be protected by restricting access to data
through access control methods or by anonymously disclosing information.
A lot of research on the topic of privacy protection has been done and in recent
years, there are a number of important review topics that analyse data confidentiali-
ty, privacy protections, proposed structures, and data privacy measures; their pros
and cons are published. Suggested methods of data privacy protection are divided
into initial methods, statistical analysis used, helpful computer-based information,
and machine learning methods.
The privacy of artificial intelligence can be violated by blocking communication
channels and accessing corporate private networks. In the area of artificial intelli-
gence, there are three major privacy risks that have taken place including privacy
issues, identity theft crimes, and unauthorized signs transmitted by criminals.
Therefore, it is important to focus on privacy while using AI networks in business
and companies should ensure that they only use independent networks rather than
third parties [12].
In order to effectively solve the data privacy protection problem, encrypted
storage of cloud data is a very outstanding solution. After encryption, the data is
stored in the server provided by cloud service provider in the form of ciphertext, and
in the meantime, the server is also required to return the data to user when the user
requires. When the user needs to use the data frequently, it requires a lot of network
bandwidth and user’s time to conduct communication with the server and realize
data encryption and decryption, which will significantly reduce the usability of
cloud computing.
In the meantime, after the encrypted data stored in cloud server has developed to
a certain scale, effective retrieval of encrypted data has become a new problem that
needs to be solved, while the traditional information retrieval technology can no long
satisfy the requirement of mass data retrieval in the cloud storage environment.
The homomorphic encryption technology is an encryption method which can
directly process the encrypted data, and it can effectively protect the security of
user’s data content, which has very broad development potential under the back-
ground of cloud storage application.
By utilizing the homomorphic encryption algorithm, it can not only ensure that
the encrypted data won’t be statistically analysed to decrypt corresponding
plaintext, but also conduct homomorphic operation (such as addition and multipli-
cation) to the ciphertext, while maintaining corresponding plaintext order of this
ciphertext during operation.
12 | Sachi Nandan Mohanty, Sirisha Potluri, V. Bhanu Prakash, B. Srinath, B. Manjunath

During the retrieval process, the used index file and keyword are both in the
form of ciphertext, and the cloud server cannot obtain any information of user data
from the retrieval results. The index file is small, which will not increase the storage
pressure of cloud server. In the meantime, the retrieval speed is fast, and it supports
retrieval of multiple keywords, which will be convenient for the users and provides
high security and strong practicability. Significance of AI in cyber safety in fig. 6.

Recurrence prevention
Initial response to alert
measures

Specification of infected Specification of infection

range source

IOC creation* IOC deployment

Fig. 6: Significance of AI in cyber safety

The following are list of privacy issues observed in cloud computing and same can
be efficiently addressed by artificial intelligence practices.
– Loss of control
– Information misuse
– Physical risk
– Identify theft
– Privacy related to internet based activates
– Privacy laws
– Network suspicious activity
– Information hacking
– Data privacy
 Cloud Security Concepts, Threats and Solutions: Artificial Intelligence Based Approach | 13

5 Security and Privacy Solutions Using Artificial

Intelligence
Artificial Intelligence is an advanced technology that helps companies manage
complex operations efficiently and improve productivity. In this generation, many
business communities are using AI-based networks to maximize organizational
performance but also face risks and threats. Security and privacy are both major
issues with AI technology that contributes to data privacy and leads to hacking
problems. The proposed study focused on security issues of AI technology and ex-
plored effective countermeasures to improve data privacy [13-15].
There are following countermeasures and strategies can be used for addressing
security threats in artificial intelligence:
– Avoid unwanted and unauthorized signals from the systems
– Only use secure networks and protect advanced systems
– Install effective security tools like firewall and data encryption
– Provide complete training to the employees related to AI based networks
– Develop and implement privacy policies
– Using anti-phishing techniques and tools

6 Proposed Solutions or Approaches

6.1 Research method
It is a large part of research-focused on research quality and questions related to the
topic of research. There are two methods used as research methods namely inclina-
tion and extraction. Import is used to provide depth details related to the research
topic while the pull method is used to give brief details on investigation. In this
study, the import method is used because of its ability to deliver practical points
related to cloud security. With the help of such method, various research questions
can be successfully resolved in order to find reliable facts related to the research
topic.

6.2 Research design

It is a useful tool that provides proper research and management of research issues
successfully related to the research. Two types of designs used in the study are
namely quality and quantity. Russell, Dewey, and Tegmark, (2015) checked that
quality is used to collect doctrinal facts however the plural is used to obtain statisti-
cal data in the study. In this study, quality construction is acceptable because of
14 | Sachi Nandan Mohanty, Sirisha Potluri, V. Bhanu Prakash, B. Srinath, B. Manjunath

ownership of the research topic and the ability to provide depth details related to
artificial intelligence protection. It uses in such a way that the quality of research is
also improved to get the right points.

6.3 Data collection

There are two parts of data collection methods including basic and secondary.
Scherer, (2015) pointed out that the primary method provides new information but
the second method provides reviewed data and facts in the study. In this study, the
second method was adopted to collect information about artificial intelligence. To
find the second data book updates various sources such as data collected in recent,
online papers websites, and books are used.

6.4 Data analysis

In this investigation, a descriptive content method is used to analyse the data being
collected because it can provide effective results and points related to the research
topic. In addition, Excel software is used to represent data collected into tables and
graphic forms. Therefore, the use of such methods has been improved research
questions resolved and managed the flow of details.
Artificial intelligence technology can be used to take action and it has already
been used to detect malware, and its functionality has already been established.
Another proof of its effectiveness is that in July 2017 Google announced that they
will integrate Cylance AI technology into VirusTotal - a Google malware detection
service. IBM is also trying to demonstrate the effectiveness of its AI technology with
intelligent search engines that carefully control the vast amount of threatening in-
formation using Watson’s cyber security.
For privacy-preserving analysis of big data confidentiality, an in-depth learning
approach is proposed. The method converts the critical part of personal information
and it has become sensitive information. To perform this process, a two-stage meth-
od is proposed namely modified sparse autoencoder types and CNN have been used
in architecture.
A modified denoising autoencoder enables data conversion and CNN sets the
path to modified data. To achieve a lower loss in data conversion, a sparsification
parameter was added to it specifically autoencoder function with Kullback split
function - Leibler.
Here, efficiency of the model is performed by the MSE (mean square error) func-
tion of the loss. To test the accuracy of the transformation process, features based
sparse denoising autoencoder algorithm fed on the installation of CNN’s advanced
 Cloud Security Concepts, Threats and Solutions: Artificial Intelligence Based Approach | 15

algorithm and reconstructed data classification. Reconstructed data classification is

included in the classes specifically black (0), white (1) and gray (2).
CNN algorithm is classified by Black class data as Gray class with details of 0.99.
Comparisons of the proposed method with an efficient autoencoder are also provid-
ed and used in Cleveland’s medical database extracted from the databases of cardi-
ovascular disease, Arrhythmia and Skoda data sets have shown that the proposed
method exceeds other standard methods.

7 Model/Data Theft Security Technology

7.1 Private separation of teacher ensembles (PATE)
This technology works by dividing training data into multiple sets in the model
training category, each training category is an independent DNN model. Independ-
ent types of DNN are used to jointly train student model by voting. This technology
ensures that student’s mind model does not disclose specific training data, thus
ensuring the confidentiality of training data.

7.2 Unique privacy protection

This technology adds audio to data or models using different privacy in the model
training phase. For example, some experts suggest how to make gradients by means
of alternative privacy measures to protect the privacy of data.

7.3 Model watermarking

This technology integrates special recognition neurons into the original model on
the model training phase. These neurons enable a special input sample to test
whether another model has been detected by stealing the original model.

8 Why is AI Technology Essential for Cybersecurity?

Cyber-attacks are increasingly varied. While there are still many indiscriminate
attacks such as malware transmission and fraudulent emails, sophisticated attacks
targeting them make it clear that targeted intentions are now a major threat. Cyber-
attacks in recent years have been hampered by their highly organized and industri-
alized structures. Personal information and various types of malware used for tar-
16 | Sachi Nandan Mohanty, Sirisha Potluri, V. Bhanu Prakash, B. Srinath, B. Manjunath

geted attacks are widely available, readily available, and inexpensive on the black
market to facilitate the identification and use of weapons are the first stages of
cyber-killing series.
Now, hackers enjoy a situation where they can launch multiple targeted attacks
in a short period of time using the tools available on the black market. This has
blurred the line between indiscriminate attacks and targeted attacks. It is now un-
derstood that normal human resources and response speed will not be able to keep
pace with the rapid changes in cyber-attacks. This is a vision that leads security
professionals to place their trust in AI technology where rapid progress is being
made.
In terms of improving efficiency and faster response, efforts are being in pro-
gress to provide a type of solution called security orchestration and automation
(SOA) for attracting attention in the United States since about 2014 and the Hexadite
of Israel. In SOA, anti-defence strategies called playbooks are described in advance.
In line with these processes, SOA can automatically perform data collection, analy-
sis, and implementation of resistance measures as shown in fig. 5 and fig. 6. Here
again, AI is expected to play a key role in dynamic planning and revitalization of
playbooks.

9 Results
The proposed study provides details about implants intelligence and their security
concerns against companies. It turns out that AI is an efficient technology that helps
companies to solve complex tasks reliably. Safety is the big deal of anxiety associat-
ed with assisted artificial intelligence technology because criminals have to obtain
sensitive information and access to computer devices of companies. A review of the
literature done and shows that ignorance is a major problem with employees as a
result they can use unauthorized networks in computer programs as well as suffers
from data breach problems.
It is found that selected search terms have been provided full details regard-
ing the security of artificial intelligence. A big part of artificial intelligence is that
it connects a person, mechanical equipment etc. so that companies can work with
a number of jobs simultaneously. Therefore, it is important for companies to use
appropriate safety tools while operating artificial intelligence at work and em-
ployees do not know about security attacks related to computer devices. Available
that there are three safe attacks that occur in the implant site.
Fraud includes malware, identity theft, and refusal to work on attack. The
proposed study provided effective combat strategies which can be used to address
privacy concerns and connected security issues with artificial intelligence tech-
nology. Data security and privacy is ensured with efficient data encryption and
 Cloud Security Concepts, Threats and Solutions: Artificial Intelligence Based Approach | 17

decryption algorithms. With AI integration, security and privacy can be height-

ened to the next level. Code for reverse cipher and Caesar cipher algorithms are
given as shown below. Flowchart for data encryption and decryption is given in
the fig. 7.

Input-Plain text

Import data

Encryption
process

Encrypt plain text

using data
encryption algorithm

Encrypted message

Import data from

encrypted message

Decrypt message
Decryption using data
process
decryption algorithm

Output- Plain text

Fig. 7: Data encryption and data decryption process

18 | Sachi Nandan Mohanty, Sirisha Potluri, V. Bhanu Prakash, B. Srinath, B. Manjunath

Listing 1: Code for reverse cipher and Caesar cipher algorithms

#Code for reverse cipher algorithm

message_var = "Demo on reverse cipher"
translated_var = " "
it_var = len (message_var) - 1
while it_var >= 0:
translated_var = translated_var + message_var [it_var]
it_var = it_var - 1
print ("Cipher text with cipher algo is: ", translated_var)
#Code for Caesar cipher
def encrypt_fun (text_var, s1):
result_var = ' '
for it_var1 in range (len (text_var)):
ch = text_var [it_var1]
if (ch.isupper ()):
result_var += chr ((ord (ch) + s1-65) %26+65)
else:
result_var += chr ((ord (ch) + s1-97) %26+97)
return result_var
text_var = "Demo on Caesar cipher"
s1 = 4
print ("Plain text: " + text_var)
print ("Shift pattern text: " + str (s1))
print ("Cipher text: " + encrypt_fun (text_var, s1))

10 Conclusion
From the above discussion, it can be concluded that security is a major problem
associated with AI technology that helps criminals commit data breaches and hack-
ing activities in companies. This paper provided information about artificial intelli-
gence and examined the security problems associated with AI technology.
It is found that the conducted literature search reviewed each and every point
related to the artificial intelligence security attacks. There are numerous attacks
occur in AI technology including malware, phishing, and DoS attack that need to be
addressed on the priority basis in order to protect private details from the criminals.
It is determined that the utilization of firewall and encryption methods can help the
companies to protect data against cybercrimes and attacks. The company should
provide training to the employees by which they can avoid the involvement of third-
party networks and unauthorized signals from the AI-based systems.
 Cloud Security Concepts, Threats and Solutions: Artificial Intelligence Based Approach | 19

With the increasing amounts of data generated by programs, traditional internal

access detection systems have worked slower. As a result, applications for artificial
intelligence are widespread although they are currently designed for population
growth. This is because people believe that they are currently doing more compre-
hensive tasks than AI goals for detecting, identifying and responding to new cyber
threats. There is the concern, however, is that AI tools themselves may be unsafe
and attackers may produce AI-powered weapons soon. This is an area of research
that needs to continue attention.
The impact of AI on human aspects of information and cyber security is there-
fore summarized in the present and future conditions. For now, state, the impact of
AI on human capital. In the future, however, the impact can go either way; that is,
people can be completely replaced by independent AI applications or individuals
and AI may be integrated to fill each other with good results. Investigators went into
future when the social and technological process is recommended as a solution and
needs to continue research in this regard. There is recognition of certain research
limitations in all aspects of AI. On the other hand, there are various AI magazines
that have not been tested.
This proposed research provided a way where readers can enhance skills in the
area of artificial intelligence and resolve the security concerns in an appropriate
manner.

11 References
[1] Mazur S, Blasch E, Chen Y, Skormin V, Mitigating cloud computing security risks using a self-
monitoring defensive scheme, Proceedings of the 2011 IEEE National Aerospace and Electron-
ics Conference (NAECON), Dayton, OH, 2011, pp. 39-45, doi: 10.1109/NAECON.2011.6183074
[2] Alabdulkarim A, Al-Rodhaan M, Tian Y, Al-Dhelaan A, A privacy-preserving algorithm for clinical
decision-support systems using random forest, Computers, Materials & Continua, 2019,
Vol.58, No.3, pp.585-601, doi:10.32604/cmc.2019.05637
[3] Chen J, Wang Y, Wang X, On-demand security architecture for cloud computing, Computer,
2012, vol. 45, no. 7, pp. 73-78, doi: 10.1109/MC.2012.120
[4] Li B, Hou B, Yu W et al., Applications of artificial intelligence in intelligent manufacturing: a
review, Frontiers Inf Technol Electronic Eng, 2017, 18, 86–96
[5] Rasim A, Ramiz A, Fargana A, Privacy-preserving deep learning algorithm for big personal data
analysis, Journal of Industrial Information Integration, 2019, Volume 15, Pages 1-14, ISSN 2452-
414X, https://doi.org/10.1016/j.jii.2019.07.002
[6] Sukhpal G, Shreshth T, Minxian X, Inderpreet S, Karan V, Lindsay D, Shikhar T, Smirnova D,
Manmeet S, Jain U, Pervaiz H, Sehgal B , Sukhwinder K, Sanjay M, Sadegh A, Harshit M,
Stankovski V, Garraghan P, Transformative effects of IoT, Blockchain and Artificial Intelligence
on cloud computing: Evolution, vision, trends and open challenges, Internet of Things, 2019,
Volume 8, 100118, ISSN 2542-6605, https://doi.org/10.1016/j.iot.2019.100118
[7] Wan J,Yang J, Zhongren W, Qingsong H, Artificial intelligence for cloud-assisted smart factory,
IEEE Access, 2018, PP. 1-1, 10.1109/ACCESS.2018.2871724
20 | Sachi Nandan Mohanty, Sirisha Potluri, V. Bhanu Prakash, B. Srinath, B. Manjunath

[8] Chen J, Dongyong Y, Data security strategy based on artificial immune algorithm for cloud
computing, Applied Mathematics & Information Sciences, 2013, 7, 149-153,
10.12785/amis/071L21
[9] Sun L, Jiang X, Ren H, Guo Y, Edge-cloud computing and artificial intelligence in internet of
medical things: architecture, technology and application, IEEE Access, 2020, vol. 8, pp.
101079-101092, doi: 10.1109/ACCESS.2020.2997831
[10] Tabrizchi H, Rafsanjani K, A survey on security challenges in cloud computing: issues, threats,
and solutions, J Supercomput, 2020, 76, 9493–9532, https://doi.org/10.1007/s11227-020-
03213-1
[11] Barona R, Anita E, A survey on data breach challenges in cloud computing security: issues and
threats, 2017 International Conference on Circuit, Power and Computing Technologies (IC-
CPCT), Kollam, 2017, pp. 1-8, doi: 10.1109/ICCPCT.2017.8074287
[12] Alia A, Al-Rodhaan M, Al-Dhelaan Y, A privacy-preserving algorithm for clinical decision-
support systems using random forest, Computers, Materials & Continua, 2019, 58, 585-601,
10.32604/cmc.2019.05637
[13] Wang T, Jiyuan Z, Chen X, Guojun W, Anfeng L, Yang L, A three-layer privacy preserving cloud
storage scheme based on computational intelligence in fog computing, IEEE Transactions on
Emerging Topics in Computational Intelligence, 2018, 2, 3-12, 10.1109/TETCI.2017.2764109
[14] Ram P, Murali S, Siva A, Privacy preservation techniques in big data analytics: a survey, J Big
Data, 2018, 5, 33, https://doi.org/10.1186/s40537-018-0141-8
[15] Muhammad N, Cloud computing: security issues and challenges, Journal of Wireless Commu-
nications, 2016, 1 (1), 10-15, https://doi.org/10.21174/jowc.v1i1.73
Shagun S Lokre, Shanmukhi Priya, Vihas Naman, Sachi Nandan
Mohanty, Sirisha Potluri
Addressing Security and Privacy in Cloud
Computing: Blockchain as a Service
Abstract: In recent times, the IT industry has been marked as a new era of software
advancement with the introduction of cloud computing technology. This technology
has become very essential for delivering infrastructure and has been adopted widely
due to its performance and availability. Cloud computing allows us to access many
applications and services over the Internet. A majority of the activities performed by
this technology revolves around different forms of data, making it an invaluable
resource. The importance and usage of data have reached new heights and like all
precious resources, even data requires security and privacy. Unfortunately, this is
an issue that concerns many cloud computing users as the data proprietors cannot
have complete control over certain aspects like the location of the data storage and
access permissions. Blockchain Technology, one of the emerging technologies can
be used to complement the issues concerning cloud users. This technology provides
features like decentralization, distribution, authenticity, immutability, trust, etc.
that can be used to face the threats of data integrity. In this chapter, we will address
the issues concerning the security and privacy of the data in the cloud model and
how the features of the blockchain technology can enhance the integrity of this
model. Although, there are certain limitations like high latency and low throughput,
decreasing the practical utility of blockchain technology

Keywords: Security, Privacy, Cloud, Decentralization, Integrity, Blockchain

Technology

||
Shagun S Lokre, Department of CSE, IcfaiTech( Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanpally, Shankarpalli Road, Hyderabad, Telangana 501203,
India, [email protected]
Shanmukhi Priya, Department of CSE, IcfaiTech( Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanpally, Shankarpalli Road, Hyderabad, Telangana 501203,
India, [email protected]
Vihas Naman, Department of CSE, IcfaiTech( Faculty of Science and Technology), ICFAI Foundation
for Higher Education, Donthanpally, Shankarpalli Road, Hyderabad, Telangana 501203, India,
[email protected]
Sachi Nandan Mohanty, Department of Computer Engineering, College of Engineering Pune,
Wellesley Rd, Shivajinagar, Pune, Maharashtra 411005, India, [email protected]
Sirisha Potluri, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana
501203, India, [email protected]

https://doi.org/10.1515/9783110732573-002
22 | Shagun S Lokre, Shanmukhi Priya, Vihas Naman, Sachi Nandan Mohanty, Sirisha Potluri

1 Introduction
Even before cloud computing was developed, it was the IT industry that faced some
tough challenges related to storing and accessing data, hosting, and data privacy to
name a few. Before the cloud, client/server computing was essentially centralized
storage where all the software, data, and control is located on the server-side. This
created a major setback for the IT industry since centralized databases rely heavily
on network connectivity, which means the slower the internet connection, the long-
er the access time for the database is needed. Although data redundancy is limited
or no, if a collection of data is lost accidentally, it is very difficult to recover it man-
ually. Earlier if a user desired to view some information or execute a program,
he/she would have to connect to the server and then obtain access to it and then do
his/her work. That’s not all, if the IT firm is huge with a large number of employees
then it becomes difficult for the company to arrange adequate hardware and soft-
ware to carry out their tasks. It is not possible to purchase software and applications
for every system and every employee. But it did not take much time to find a solu-
tion that changed the face of the IT industry, “Cloud Computing” [1-2]. Cloud com-
puting in simple terms means that the services provided of computer resources are
controlled and managed by the cloud provider rather than the end-user. These re-
sources include everything from browser-based web applications, third party data
storage devices, or third-party servers used to support an industry, research, or
personal project’s computer infrastructure.
Before the widespread use of cloud computing, businesses and general comput-
er users usually had to purchase and maintain the software and hardware they
wanted to use. Based on the increasing availability of cloud-based apps, storage
services, and machines, firms and consumers have access now as Internet services
to a wealth of on-demand computing resources [3]. If on-the-job software and hard-
ware are shifted to a remote networked and distributed resource, cloud users are no
longer required to invest the work, equity, or expertise needed to buy and maintain
the data. This unprecedented access to computer resources has resulted in a new
group of cloud-based companies and changed IT practices across industries and
turned many computer-aided practices into everyday ones. Many organizations
started to use the cloud to store records, supply corporate applications, and deploy
services and goods digitally. These kinds of cloud-based technologies and adoptions
are unique to the field. In the education world, educators and scholars use cloud-
based teaching and research applications [4]. As they say, nothing is perfect in this
world, cloud computing has few drawbacks which need to be taken into considera-
tion before moving ahead. In cloud computing, confidentiality and privacy are
amongst the most questionable. By the use of cloud storage infrastructure, we are
completely committed to providing cloud-based application protection and confi-
dentiality of the data. This means that if we encounter any problem, we cannot
 Addressing Security and Privacy in Cloud Computing: Blockchain as a Service | 23

claim damages to the server for data errors or any kind of mistakes in the data. An-
other major disadvantage in cloud computing is its vulnerability in the event of an
attack. There are several concerns about cloud computing since cloud computing is
online, any aspect of cloud computing can be exposed to a wide range of servers
allowing data attacks and server storage activities. In this chapter, we are going to
propose how Blockchain Technology can be useful to make the whole system decen-
tralized, tamper-proof, and transparent [5].

2 Literature Survey
As of now, blockchain technology faces core obstacles related to scalability, regula-
tory limits, identity registration, consumer protection, laws and regulations, and
compliance requirements. This chapter is written based on the literature of the pre-
vious research, combines blockchain technology and the concept of cloud to create
a secure cloud system with selected access permissions [6]. This chapter consists of
a detailed study of the combination of cloud systems and blockchain technology.
With the current progress of science and technology, the need for secure and fast
systems has become the top priority for any information retrieval. The current uses
of blockchain include assistance to an IoT system and cloud computing system
assisting blockchain technology to make a home environment more efficient, se-
cure, and faster in terms of any processes, this uses Cloud Service Provider [7], Edge
Service Provider, and a Lightweight client; each of these modules assists in the stor-
age and usage of data.
Another application using blockchain is secure network computing for light-
weight clients, this application is used for mobile devices that do not have enough
computing power for heavy usage [8]. This application uses mobile resource man-
agement that connects nearby mobile phones using Bluetooth technology, with the
combined resources of the phones nearby we use Secure Authentication Manage-
ment human-centric Scheme (SAMS) to authenticate mobile devices using block-
chain, with all the devices giving authentication data falsification can’t take place
[9]. The major usage of blockchain technology is the medical sector, the medical
history of any patient is really important for the diagnosis of specific illnesses [10].
The easier it is to store and securely share this data as confidentiality is of utmost
importance, the easier it is to diagnose and cure any illness. Using a decentralized
application, we will upload all the medical history to the blockchain, allowing the
patient to grant access to any specific doctor to check and add the diagnosis to the
existing data. This concept is known as Electronic medical records. We have also
used multiple survey papers related to the security and privacy issues present in
cloud computing. There are multiple types of cloud services such as Software, Plat-
form, and Infrastructure as a service [11].
24 | Shagun S Lokre, Shanmukhi Priya, Vihas Naman, Sachi Nandan Mohanty, Sirisha Potluri

Each of these models has complexities and security issues of its own. The main
challenges being confidentiality, Integrity, availability of data [12]. The supporting
concept for cloud computing is virtualization, the process which allows easy up-
loading of data onto any cloud service. There are multiple challenges present in
technology related to cloud computing using blockchain. From the blockchain per-
spective, there are scalability issues. From the cloud computing point of view, there
are security problems and financial problems since cloud services can be expensive
[13].

3 Cloud Computing Concepts

Before moving ahead with the chapter, one might come across a few questions such
as what exactly cloud is? Where is the cloud? Are we currently in the cloud? Cloud
computing, in the most simplistic words, means the data and programs are stored
and accessible through the Internet rather than the hard disc of our computer. In
the end, the term “cloud” is just an internet metaphor. Cloud-based computing en-
courages files to be saved in a remote archive instead of on a private hard disc or
local storage device. Although an electronic computer has access to the Internet, the
electronic device has access to the data and the software application. Cloud storage
is called as such since the knowledge being viewed can be located directly in a
cloud or a virtual environment. Companies that have cloud storage allow users to
save files and software on remote servers and then access all data over the Internet.
This ensures that the user is not forced to be in a certain position to obtain access to
it, enabling the user to operate remotely. Cloud computing eliminates all of the hard
lifting involved in crunching and manipulating data from the laptop we hold or sit
and operate on. It also transfers the entire job to huge cyberspace computing clus-
ters. The Internet is evolving into a cloud, such that our files and jobs can be viewed
from every device in the world, from which we bind to the Internet. Cloud storage
platforms provide a range of features for consumers, such as Email, memory, data
backup, development of and checking software, data processing, audio and video
streaming, distribution of on-demand devices [14].
The services provided by cloud computing is categorized into 3 service models
as shown in the fig. 1:
– SaaS: Software as a Service.
– PaaS: Platform as a Service.
– IaaS: Infrastructure as a Service

SaaS: NIST describes this service in cloud as being able to access the provider’s
cloud computing applications. Applications can be accessed from separate client
devices by either a thin client interface such as a web browser or a software inter-
 Addressing Security and Privacy in Cloud Computing: Blockchain as a Service | 25

face. With the likely exception of minimum user specific configuration require-
ments, the user does not track or manage the cloud resources including network
processors, operating systems, memory and even individual software functionality.

SaaS
CRM, Office suit, Email, Games

PaaS
Cloud Client
Applications, Development, Deployment
Web browser, Mobile App, etc.
tools

IaaS
Applications, Development, Deployment
tools

Fig. 1: Categories of cloud computing

Business Apps (Business Users)

IDE (Developers)
Platform as a service
Integrated Development Environment

Data security (Data security analyst)

Backup/ Recovery
Scalable infrastructure
Application hosting

Fig. 2: Platform as a service

26 | Shagun S Lokre, Shanmukhi Priya, Vihas Naman, Sachi Nandan Mohanty, Sirisha Potluri

Users gain access to application applications and databases as a service (SaaS)

model. Infrastructure and systems that run software are operated by cloud provid-
ers. SaaS is often referred to as “on-demand software” and is typically priced on a
pay-per-use basis or with a monthly fee.
The use of SaaS has proven to be advantageous in terms of scalability, reliabil-
ity, and performance.
– Modest software tools
– Effective use of software license
– Centralized data and monitoring
– Responsibilities of the network operated by the provider
– Multi-stakeholder solutions

Block

Block header

Block Information

Block version
Merkle tree root hash
Nounce
Previous block hash
Timestamp
Difficulty target

List of transactions

Fig. 3: Block Structure in a blockchain

 Addressing Security and Privacy in Cloud Computing: Blockchain as a Service | 27

PaaS: As per NIST, the platform as a service definition in cloud computing is de-
scribed as: The functionality offered to the user is to execute consumer-created or
purchased software to the cloud infrastructure by writing a set of codes, using li-
braries, APIs, and resources funded by the vendor. The user neither operates nor
monitors the underlying cloud resources, including the network, servers, operating
systems, or storage, but has control over the software installed and likely configura-
tion settings for the application-hosting environment.
PaaS vendors provide a platform for creating applications as shown in fig. 2. The
vendor usually creates a toolkit, implementation specifications and channels for
distribution and payment. In PaaS models, cloud vendors include a computing in-
frastructure, usually providing an operating system, a programming-language exe-
cution environment, a database, and a web server.
The benefits of using PaaS include:
– Higher administrative overheads
– Adjustable options
– More modern software for the framework

IaaS: Infrastructure as a Service (IaaS) is an Instant Computing Infrastructure, de-

livered and operated over the Internet. It is one of four types of cloud computing,
along with software as a service (SaaS), application as a service (PaaS), and server
less services.
IaaS easily scales up and down with demand, allowing you to pay for what
you’re using. It lets you escape the cost and difficulty of owning and maintaining
your physical servers and other data centre resources. Each resource is provided as
a separate service component and you only need to rent a single resource for as long
as you need it. The cloud storage vendor, such as Azure, operates the platform when
buying, downloading, configuring, and maintaining your software—operating sys-
tems, middleware, and programs [15].

4 Consensus Mechanisms
Cloud computing provides multiple advantages when it comes to the business and
IT sectors. It allows us to create what is indeed a virtual workplace, so that we can
communicate everywhere at all times. With the number of web-enabled gadgets
used in today’s corporate world (ex. smartphones, tablets), access to your data is
now simpler. Moving on to cloud storage will decrease the IT systems administra-
tion and maintenance costs. It will reduce the expenses by leveraging the services of
the cloud computing platform rather than buy costly systems and appliances for our
business. Majority of the companies with more than 1000 employees do not have the
same IT requirements as a start-up [16]. Cloud usage is a great choice as it enables
28 | Shagun S Lokre, Shanmukhi Priya, Vihas Naman, Sachi Nandan Mohanty, Sirisha Potluri

companies to scale up/down their IT divisions efficiently—and quickly—according

to market requirements.
Cloud-based applications are suitable for organizations that have rising or
evolving demands for bandwidth. You can quickly expand the capability of your
cloud without spending on physical resources as your company demands in-crease.
This degree of versatility will offer a true edge over competing companies using
cloud computing. That’s when scalability comes into the picture. Scalability mini-
mizes the risks involved with internal maintenance and organizational challenges.
With professional strategies and zero upfront investment, you have high-
performance tools at your side. The biggest value of the cloud is undoubtedly scala-
bility [17].

5 Blockchain Technology Concepts

In layman’s terms, a blockchain is a chain of blocks that contains data in the form of
transactions. Technically, a blockchain is a distributed, decentralized, and trusted
ledger or data-sharing platform that can be utilized by participants who do not trust
each other. A blockchain is said to be distributed as every participant in the network
executes a job collectively and it is said to be decentralized as there are multiple
points of coordination, with no single point of failures [18].
In the earlier days, the usage of traditional ledgers resulted in central authority
as only one entity would control the entries made into the ledger. But with the in-
troduction of blockchain technology, each participant in the network would have a
copy of the ledger and no single participant holds the power to make an entry sole-
ly. All the participants follow a consensus mechanism to decide the addition of an
entry into the ledger. The most common consensus mechanisms are listed below.

5.1 Consensus mechanisms

A consensus mechanism is an algorithm for fault tolerance that is used by the
blockchain technology to accomplish the agreement requirements on an individual
data value or a single state of the network amid the participants [19].
– Proof of Work (PoW): This is the most commonly used algorithm these days.
This algorithm requires the participants of the network to solve a challenging
computational puzzle to create new blocks. The mechanism of solving the puz-
zle is called “mining” and the nodes that carry out this process are called “min-
ers”. The miners who solve this puzzle will get an incentive, known as a block
reward. The probability of mining a block depends on the work done by the
miner. This work consumes physical resources like CPU power and time.
 Addressing Security and Privacy in Cloud Computing: Blockchain as a Service | 29

– Proof of Stake (PoS): This algorithm requires a miner to acquire a sufficient

stake to mine a block. While executing an attack is expensive, the incentive for
the attack is also reduced as the attacker needs to own a high stake. There-fore,
an attack will have more effect on the attacker. This process provides in-creased
security and is power efficient but fails to keep the system completely decentral-
ized.
– Proof of Burn (PoB): This algorithm requires the miner to burn some wealth to
mine a new block. The miners should show proof that they have burned some
wealth by sending coins to a verifiably un-spendable address. This process is
expensive like PoW but no external resources are used except the miner’s
wealth. PoB consumes virtual or digital resources and is power efficient too.
PoB works by burning PoW mined cryptocurrencies.

5.2 Structure of blockchain

All blockchain structures are generally categorized into three categories:
– Public blockchain architecture: In this architecture, any participant can enter
and exit the blockchain network as per their will, without any permissions. All
the willing participants have access to the data in the network as long as they
are in the network. Ex: Bitcoin, Ethereum, etc.
– Private blockchain architecture: In this architecture, each participant re-quires
permission to enter and exit the blockchain network. The participants who enter
the blockchain can be restricted to certain data as well. This system is controlled
by the users of a specific organization. Ex: Hyperledger Fabric.
– Consortium blockchain structure: In this architecture, the blockchain net-work
is built by a set of organizations. This architecture is neither completely public
nor completely private. All the permissions are controlled by the preliminary as-
signed users. Ex: Quorum, Corda, etc.

The tab.1 gives an outline of the comparison of the three blockchain structures:

Tab. 1: Blockchain structure comparison

Property Public Blockchain Private Blockchain Consortium Blockchain

Accessibility Anyone Single organization Multiple organizations

Who can join? Anyone Permissioned and Permissioned and
known identities known identities
Consensus mechanism Pow/Pos Voting or multi- Voting or multi-
party consensus party consensus
algorithm algorithm
30 | Shagun S Lokre, Shanmukhi Priya, Vihas Naman, Sachi Nandan Mohanty, Sirisha Potluri

Property Public Blockchain Private Blockchain Consortium Blockchain

Consensus All miners Within one organiza- A selected set of nodes

determination tion
Consensus process Permission less Permissioned Permissioned
Read permission Public Public or restricted Public or restricted
Immutability level Almost impossible to Could be tampered Could be tampered
tamper
Efficient usage of Low High High
resources
Centralization No Yes High
Transaction speed Slow Lighter and faster Lighter and faster

Though a blockchain is divided into three categories, all blockchains consist of

blocks and all blocks follow a similar structure. A block is like a data structure that
accumulates the transactions that would be included in the blockchain [20]. Every
block is divided into a block header that contains metadata and a list of transactions
as shown in fig. 3. The first block in the blockchain is called the genesis block. A
block header consists of:
– Block version: This is a version number that is used to track software updates.
Its size is 4 bytes.
– Previous block hash: This is a reference to the hash of the parent/previous block
in the current chain. Its size is 32 bytes.
– Merkle tree root hash: This is the hash of the root of the Merkle tree of this
block’s transactions. Its size is 32 bytes.
– Timestamp: This is the approximate time required to create a block. Its size is 4
bytes.
– Nonce: This is a counter used for the proof-of-work algorithm. Its size is 4 bytes.
– Difficulty target: This is the proof-of-work algorithm difficulty target for this
block. Its size is 4 bytes.

The mining process requires the nonce, the difficulty target, and the timestamp.

5.3 Workflow of blockchain

Fig. 4 depicts the workflow of a blockchain that utilizes the proof-of-work consensus
mechanism.
 Addressing Security and Privacy in Cloud Computing: Blockchain as a Service | 31

A transaction is initiated
and it is broadcasted to
the other participants in
the network

The other participants The other participants

verify the block, append validate the transaction
it to the current chain, and record the
and add it to their copy transaction in to the
of the local ledger newly mined block

The newly mined block is

broadcasted in to the
network

Fig. 4: The workflow of a blockchain

Firstly, a transaction is initiated by a participant and it is broadcasted to the other

participants in the network. The participants who get this transaction use the digital
signature to check the authentication of the transaction. After the verification, the
transaction is attached to the list of authentic transactions in the participant’s re-
spective ledgers. To register the authentic transactions, the participants in the net-
work work to mine the new block, i.e., finding the nonce of the block. Once a partic-
ipant finds a valid nonce, they can mine the block that contains the concerned
transaction. The other participants then check the trans-actions in the received
block by analysing the Merkle root, and once the transactions in the newly mined
block are authenticated and have not been tampered with, the new block is append-
ed to the local replica of the blockchain. Once this entire procedure has been com-
pleted, the newly mined block can be appended to the current blockchain [21]. Sam-
ple code for blockchain as a service to declare the data of transactions and
blockchain mining is given as below.
32 | Shagun S Lokre, Shanmukhi Priya, Vihas Naman, Sachi Nandan Mohanty, Sirisha Potluri

Listing 1: Blockchain as a service to declare the data of transactions and blockchain mining

#Transactions declaration
def get_transdata(self, sender, receiver, amount):
self.current_data.append({
'sender': sender,
'receiver': receiver,
'amount': amount
})
return True
#Blockchain mining
def blockchain_mining(self, details_miner):
self.get_data(
sender="0",
receiver=details_miner,
quantity=1,
)
last_block = self.latest_block
last_proof_number = last_block.proof_number
proof_number = self.proof_of_work(last_proof_number)
last_hash = last_block.compute_hash
block = self.build_block(proof_number, last_hash)
return vars(block)

5.4 Features of blockchain

A few important features of the blockchain are listed below [22-25]:
– Decentralization: According to this feature, all the transactions are registered by
all the participants in the network and each participant maintains a local copy
of the ledger, where the recorded transactions are saved. In this way, the ledger
is shielded from a central authority and a single point of failure.
– Trustable: According to this feature, the participants entering the network do
not need to prove their identity as the participants need not depend on any par-
ticular participant for the validation of any transaction. There is no requirement
of any third-party as the consensus algorithm is used for validating the transac-
tions.
– Immutability: According to this feature, the transactions recorded into the
blockchain can never be manipulated or tampered with. The cryptographic
principles are utilized to make the blockchain and its contents immutable.
– Non-repudiation: According to this feature, no participant can deny mak-
ing/receiving a transaction once the transaction is done. This is because every
 Addressing Security and Privacy in Cloud Computing: Blockchain as a Service | 33

transaction is cryptographically signed with a private key before broadcasting it

to the network. This private key is known only to its owner, therefore making it
impossible for an initiator to deny a transaction made by him/her.
– Transparency: According to this feature, every participant can access all the
transactions recorded in the blockchain and verify all the transactions.
– Traceability: According to this feature, each transaction and block can be traced
and verified easily with the help of the timestamp that is attached to the block
header.

5.5 Potential attacks on blockchain

Like we know that every coin has two sides, even blockchain technology has certain
vulnerabilities, leading to attacks on the blockchain. Some potential attacks are
listed below:
– Sybil attack: An attacker floods the entire network with an enormous number of
nodes with anonymous identities. In this way, the attacker can have control of a
majority of the network and can try to dominate the network. Though the nodes
seem to be unrelated, they are operated by a single partici-pant. This attack fo-
cuses on the entire network allowing the attacker to create a fork in the ledger
making double-spending possible.
– Selfish mining attack: Generally, the longest chain in the network is con-sidered
to be the true and latest version of the ledger. Therefore, if a selfish miner builds
blocks on top of an existing chain in stealth mode, he/she will be able to pub-
lish this fake chain and everyone would accept this chain as it’s the longest.
This provides a small window to the attacker to perform double-spending.
– 51% attack: This attack occurs when a miner or a group of miners controls 51%
or more of the mining power in a blockchain network. This is difficult to happen
in large networks but it is highly possible in small networks.
– Denial of Service (DoS) attack: This happens when malicious participants com-
pletely absorb and keep the resources busy in verifying and transmitting the
blocks and transactions. These malicious participants can flood the network
with several transaction initiations to other participants, disabling the trans-
mission and verification of transactions from other participants.

6 Cloud Security Concerns

The biggest problem today is the challenge to cloud protection. Malicious or unin-
tentional acts can cause harm to an organisation at several levels. This is true in
both cloud and non-cloud settings, but as the complexity of software and services
34 | Shagun S Lokre, Shanmukhi Priya, Vihas Naman, Sachi Nandan Mohanty, Sirisha Potluri

grows, security vulnerabilities also increase. The cloud attack surface is higher than
the conventional service models, as the associated modules have several different
endpoints and protocols to handle in different ways. A variety of methods are need-
ed to recognise and resolve both identified and new threats.
Distributed-Denial-of-Service (DDoS) attacks are only one common security
threat in cloud computing. Providers typically provide a variety of architectural
solutions to configure a protected infrastructure to avoid these kinds of attacks,
such as traffic separation (the ability to isolate clusters of virtual machines in groups
of virtual different networks) or access control lists to establish guide-lines that
specify multiple-level component permissions.
The second most significant problem for conventional cloud paradigms is Data
Management. One part of data management that needs to be discussed is the con-
sistency of the data. This means that all cloud users can see the same data simulta-
neously. This is no easy job, since hardware (data resources) are paired with the
transactional status of operations in line with physical network constraints which
can affect continuity by linking the user.
The third major concern is Resilience and Cloud compatibility. Resilience is the
ability to manage errors gracefully and to restore the whole system. This is a mas-
sive problem for services and applications where modules fight for resources and
depend on other internal or external components/services that malfunction or may
rely on faulty software. Planning the manner in which those errors are observed,
logged, repaired and restored re-quires not only developers but all teams as part of a
cloud strategy.
To support, applications are available to replicate random problems, from
hardware problems to major external threats, including failed implementations or
suspicious device behaviour [26-27].

7 Blockchain to the Rescue

There are challenges and limitations in cloud computing apart from the ones that
are mentioned above, some of these limitations can be removed or be mitigated
using Blockchain Technology. The major challenges in cloud computing are security
issues, cost management, lack of resources, governance or control of data, manag-
ing multiple clouds, building a private cloud, limited features, data mobility and a
reliable internet connection.
 Addressing Security and Privacy in Cloud Computing: Blockchain as a Service | 35

Cloud user layer: cloud users requests/responds to/from

the cloud automation layer

Cloud automation layer: Inspect, organize, execute,

record and comprehension of blocks

Blockchain management layer: Open shared ledger,

consensus, shared contract and cryptography

Secure data storage layer: Blockchain cloud storage

Fig. 5: Blockchain as a service (BaaS)

The major part of these problems can be resolved using blockchain technology.
Blockchain as a service is shown in fig. 5. The characteristics of blockchain such as
decentralization, Immutability, consensus mechanism, enhance security. The fol-
lowing problems can be solved using blockchain technology.
36 | Shagun S Lokre, Shanmukhi Priya, Vihas Naman, Sachi Nandan Mohanty, Sirisha Potluri

To solve security and privacy issues that are prevailing in cloud computing,
blockchain uses secure exchange of transaction framework as shown in the fig. 6.

Input- A transaction
Emission of
request
transaction

A block that
represented the Packaging of
transaction
trasaction is created

The block is sent to

every node in the Emission of block
network

Validation of
Nodes validate the
transaction
transaction

Nodes receive a
Rewarding of reward for the proof
transaction
of work

The block is added

Addition of block to the existing
blockchain

Reception of
transaction Output- Transaction

Fig. 6: Secure exchange of transaction framework in blockchain

 Addressing Security and Privacy in Cloud Computing: Blockchain as a Service | 37

7.1 Encryption
Although cloud services providers implement the best security standards and great
certifications. Storing sensitive information will always come with the risk of mali-
cious users and hackers trying to get the information. For example, the hacking of
AWS EC2 console at Code Space, which led to data deletion and eventually the clos-
ing of the company. In Blockchain-as-a-Solution (BaaS), data is broken down into
many encrypted segments, which are connected by a hashing feature.
These protected portions are split throughout the network and are decentralised
in each segment. Security provisions are high such as account ledgers, pub-
lic/private key encryption and haze blocks. It provides accurate and stable hacker
protection. Due to sophisticated 256-bit encryption even experienced hackers can-
not decrypt info.

7.2 Cost management

Maintaining a server or a database for storage is not cost effective, blockchain is
decentralized which is secure and not in a single place or on a single server which
reduces the cost by a marginal amount and it also decreases the storage needs for
any firm. The security is not a problem as the blockchain has high security
measures. As the load for retrieving any data is spread across multiple systems it is
easier and faster to retrieve any files.

7.3 Lack of resources

There is a need for an expanding storage, in this scenario the use of blockchain is
highly advantageous. We can use the concept of distributed ledger technology to
distribute the load across multiple systems. If we are using a cloud or a centralized
system it is costly to improve the speed of retrieving data.

7.4 Governance or control of data

There is lesser control of the transfer of data, it is difficult to keep track of who is
accessing the data in a cloud environment whereas in the blockchain technology we
can track each request and pull of data from the on-chain data or the data on the
blockchain. Each query is stored on the blockchain as a transaction and it is irre-
versible. This makes it easier to grant permission for access of data and tracking of
data queries is systematically filtered. Suppose there is a corruption of data on the
chain, the system can be put offline and the copies of the data on the other blocks
can be retrieved.
38 | Shagun S Lokre, Shanmukhi Priya, Vihas Naman, Sachi Nandan Mohanty, Sirisha Potluri

7.5 Managing multiple clouds or servers

With the ever increasing need to store any sort of data, we need an expanding stor-
age. This is too costly to manage. The next solution is to manage multiple servers or
cloud environments. This makes it really difficult to track and filter the data. In
blockchain we can use multiple systems and solve this issue. We don’t require mul-
tiple servers as the data on blockchain is stored in a decentralized manner.

7.6 Building a private cloud

If the data or any information is highly confidential then there is a need for a private
cloud. This is not cost effective and costly; this can be solved using a private block-
chain. Using a private blockchain has multiple advantages such as select granting
of access and secure data transfer. Building a private blockchain is much easier than
having a private cloud. Maintaining the storage modules and the server setup is a
tedious and high maintenance work.

7.7 Limited features

Not all the cloud services were made with the same features. If there is a need for
changing any characteristic on the cloud, migration of all the data is an extremely
long and painstaking process. Using blockchain we can decide the initial set of
characteristics and use them. As blockchain technology is constantly evolving the
security measures and any other features can be updated whenever required.

7.8 Reliable internet connection

It is highly important to have a fast and strong connection using the internet or the
intranet. The speed should be maintained from the retriever’s side and the server-
side. As the data is stored across multiple systems in a blockchain environment. A
temporary downtime for a particular system will mean that the same data can be
retrieved from a different block from the chain of blocks. This is one of the biggest
problems in a cloud environment which can be solved using blockchain technology.

8 Conclusion
Cloud computing has given us multiple systems to store and manipulate data. We can
definitely use blockchain to improve security systems and data sharing as compared to
 Addressing Security and Privacy in Cloud Computing: Blockchain as a Service | 39

the system that is present now. In this chapter, we addressed blockchain technology and
associated key innovations and looked at the pattern of studies to date to explore more
fields to be explored. Various existing concerns for the usage of blockchain in the cloud
computing world should be taken into consideration. Blockchain is causing many prob-
lems right now, such as transaction stability, wallet, and applications, and multiple
types of research have been conducted to fix the above-mentioned issues. The anony-
mous information of the user should be maintained in a secure manner by using block-
chain in the cloud computing environment, and user information should be deleted
when the user is withdrawing or discontinuing the usage of the service. If the infor-
mation of the user is not discarded, it can be used to figure out who the user is and the
identity of the user is exposed, the information retrieved can be used for malpractice.
Cloud infrastructure is here to remain, but be careful—Traditional methods will
not be enough to solve the demands of current cloud workloads. Cloud storage, data
management, and resource compatibility are real issues that need to be taken into
consideration as part of a modern cloud strategy.

9 References
[1] Smith F, Waterman S, Identification of common molecular subsequences, J. Mol. Biol, 1981,
147, 195–197
[2] May P, Ehrlich H, Steinke T, ZIB structure prediction pipeline: composing a complex biological
workflow through web services, Nagel W, Walter W, Lehner W, Euro-Par 2006 Parallel Pro-
cessing, LNCS, 2006, vol. 4128, pp. 1148–1158, Springer, Heidelberg
[3] Toby V, Anthony V, Elsenpeter R, Cloud computing, A Practical Approach (1st. ed.), McGraw-
Hill, Inc., USA, 2009
[4] Sabahi F, Cloud computing security threats and responses, 2011 IEEE 3rd International Confer-
ence on Communication Software and Networks, 2011, pp. 245-249, doi:
10.1109/ICCSN.2011.6014715
[5] Jin P, Jong P, Blockchain security in cloud computing: use cases, challenges, and solutions,
Symmetry, 2017, 9, no. 8: 164
[6] Ravishankar B, Kulkarni P, Vishnudas M V, Blockchain-based database to ensure data integrity
in cloud computing environments, 2020 International Conference on Mainstreaming Block
Chain Implementation (ICOMBI), Bengaluru, India, 2020, pp. 1-4, doi:
10.23919/ICOMBI48604.2020.9203500
[7] Saurabh S, In-ho R, Weizhi M, Maninder K, Cho G, SH-BlockCC: A secure and efficient Internet
of things smart home architecture based on cloud computing and blockchain technology, In-
ternational Journal of Distributed Sensor Networks, 2019, 15(4), 155014771984415,
10.1177/1550147719844159
[8] Yang X, Guojun W, Jidian Y, Ju R, Yaoxue Z, Cheng Z, Towards secure network computing ser-
vices for lightweight clients using blockchain, Wireless Communications and Mobile Compu-
ting, 2018, 1-12, 10.1155/2018/2051693
[9] Hyun-Woo K, Young-Sik J, Secure authentication-management human-centric scheme for trust-
ing personal resource information on mobile cloud computing with blockchain, Human-centric
Computing and Information Sciences, 2018, 8 (1), 10.1186/s13673-018-0136-7
40 | Shagun S Lokre, Shanmukhi Priya, Vihas Naman, Sachi Nandan Mohanty, Sirisha Potluri

[10] Yue X, Wang H, Jin D, Mingqiang L, Jiang W, Healthcare data gateways: found healthcare intel-
ligence on blockchain with novel privacy risk control, Journal of medical systems, 2016,
40(10),218, 10.1007/s10916-016-0574-6
[11] Simanta S, Application of blockchain in cloud computing, International Journal of Innovative
Technology and Exploring Engineering (IJITEE), 2019, ISSN: 2278-3075, Volume-8 Issue-12
[12] Rajeswari S, Kalaiselvi R, Survey of data and storage security in cloud computing, 2017 IEEE
International Conference on Circuits and Systems (ICCS), 2017, pp. 76-81, doi:
10.1109/ICCS1.2017.8325966
[13] Ashok G, Siddiqui S, Alam S, Mohammed S, Cloud computing security using blockchain, JETIR,
2019, Volume 6, Issue 6, 791-794
[14] Jadeja Y, Modi K, Cloud computing - concepts, architecture and challenges, 2012 International
Conference on Computing, Electronics and Electrical Technologies (ICCEET), Kumaracoil, 2012,
pp. 877-880, doi: 10.1109/ICCEET.2012.6203873
[15] Gibson J, Rondeau R, Eveleigh D, Tan Q, Benefits and challenges of three cloud computing
service models, 2012 Fourth International Conference on Computational Aspects of Social Net-
works (CASoN), Sao Carlos, 2012, pp. 198-205, doi: 10.1109/CASoN.2012.6412402
[16] Kawanami Y, Kato H, Yamaguchi H, Tanimura M, Tagaya Y, Mechanism and control of cloud
cavitation, J. Fluids Eng, Dec 1997, 119(4), 788-794 (7 pages),
https://doi.org/10.1115/1.2819499
[17] Li X, Zhou L, Shi Y, Guo Y, A trusted computing environment model in cloud architecture, 2010
International Conference on Machine Learning and Cybernetics, Qingdao, China, 2010, pp.
2843-2848, doi: 10.1109/ICMLC.2010.5580769
[18] Liang Y, Blockchain for dynamic spectrum management, Dynamic Spectrum Management,
Signals and Communication Technology, Springer, Singapore, 2020, pp 121-146,
https://doi.org/10.1007/978-981-15-0776-2_5
[19] Nguyen C, Hoang D, Nguyen D, Niyato D, Nguyen H, Dutkiewicz E, Proof-of stake consensus
mechanisms for future blockchain networks: fundamentals, applications and opportunities,
IEEE Access, 2019, 7, 85727–85745
[20] Liu Z, Luong N, Wang W, Niyato D, Wang P, Liang Y, Kim D, A survey on blockchain: a game
theoretical perspective, IEEE Access, 2019, 7, 47615–47643
[21] Perboli G, Musso S, Rosano M, Blockchain in logistics and supply chain: a lean approach for
designing real-world use cases, IEEE Access, 2018, 6, 62018–66202
[22] Konstantinidis I, Siaminos G, Timplalexis C, Zervas P, Peristeras V, Decker S, Blockchain for
business applications: a systematic literature review, Abramowicz W, Paschke A, Business In-
formation Systems, BIS 2018, Lecture Notes in Business Information Processing, 2018, vol
320, Springer, Cham, https://doi.org/10.1007/978-3-319-93931-5_28
[23] Umesh B, Sudeep T, Karan P, Pimal K, Sudhanshu T, Neeraj K, Mamoun A, Blockchain for In-
dustry 4.0: a comprehensive review, 2020, IEEE Access, vol. 8, pp. 79764-79800, doi:
10.1109/ACCESS.2020.2988579
[24] Kshetri N, Can blockchain strengthen the internet of things?, IT Professional, 2017, vol. 19, no.
4, pp. 68-72, doi: 10.1109/MITP.2017.3051335
[25] Arora D, Gautham S, Gupta H, Bhushan B, Blockchain-based security solutions to preserve
data privacy and integrity, 2019 International Conference on Computing, Communication, and
Intelligent Systems (ICCCIS), 2019, pp. 468-472, doi: 10.1109/ICCCIS48478.2019.8974503
[26] Houshyar H, Rashid M, Fakhrul A, Demidenko S, Multi-layer blockchain-based security archi-
tecture for internet of things, Sensors 2021, 2021, 21(3), 772,
https://doi.org/10.3390/s21030772
[27] Rui Z, Rui X, Ling L, Security and privacy on blockchain, ACM Comput. Surv, 2019, 52, 3, Article
51, 34 pages, DOI:https://doi.org/10.1145/3316481
Sirisha Potluri, Sachi Nandan Mohanty, T. Kundana, D. Abhinav, P.
Sushrutha
Security and Privacy Preservation Model to
Mitigate DDoS Attacks in Cloud
Abstract: Denial-of-service (DoS) attack occurrences are among the key security chal-
lenges within the rising cloud computing models. Currently, a number of different
types of DoS attacks are conducted against the various cloud services and resources,
which target their availability, service level agreements SLAs, and performance. A
successful DDoS attack would possibly lead to service degradation or complete out-
age. DDoS attacks will reduce the network re-sources which may result in bandwidth
depletion. The main objective of DDoS attacks is to prevent certain users from access-
ing the services in the cloud. In this research, we propose a privacy-preserving cross-
domain at-tacks detection scheme for SDNs. Predis may be a combination of perturba-
tion encryption and encoding to guard the privacy and uses a computationally simple
and efficient machine learning algorithm k-Nearest Neighbour (KNN) for its detection.
Here we also improve the KNN to achieve much better efficiency. Through extensive
simulations and theoretical analysis. The DEMO of this algorithm is capable to realize
the detection process efficient and accurate attack detection by keeping sensitive
information secure. Virtualization enables simultaneous sharing of resources from a
single server to some applications or services, this induces a certain amount of threats
and security issues for the cloud. This module focuses on identifying challenges relat-
ed to cloud security and tries to introduce ways in resolving these issues

Keywords: Denial-of-Service, KNN, Software Defined Networks, Cryptography,

Virtualization, Cloud Security, Cloud Privacy

||
Sirisha Potluri, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana
501203, India, [email protected]
Sachi Nandan Mohanty, Department of Computer Engineering, College of Engineering Pune,
Wellesley Rd, Shivajinagar, Pune, Maharashtra 411005, India, [email protected]
T. Kundana, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI Foundation for
Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana 501203, India,
[email protected]
D. Abhinav, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI Foundation for
Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana 501203, India,
[email protected]
P. Sushrutha, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI Foundation
for Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana 501203, India,
[email protected]

https://doi.org/10.1515/9783110732573-003
42 | Sirisha Potluri, Sachi Nandan Mohanty, T. Kundana, D. Abhinav, P. Sushrutha

1 Introduction
The practice of employing a network of remote servers hosted on the web to store,
manage, and process data, instead of a local server or a private computer is called
cloud computing. It is a pay-for-use model. Cloud computing can be considered as
the future of the emerging technologies because of the combination of emerging
technologies like virtualization and service oriented architecture. Based on the de-
ployment models there are various types of clouds [1].
– Public/Internet Cloud
– Private/Enterprise Cloud
– Hybrid/Mixed Cloud
– Community Cloud

Public cloud is designed for a substantial set of public clients, Private clouds ser-
vices are available only for the cloud owners. The combination of two or more
clouds can be considered as hybrid cloud where in the two clouds can be either
public or community or private clouds, broadly community clouds services are
shared between several organizations.
Cloud provides services using various models like Software as a Service, Plat-
form as a Service and Infrastructure as a Service.
– SaaS: It provides user access to software applications. These software applications
which are present in the cloud are used for an enormous range of tasks.
– PaaS: It provides all the hardware and software components that are required to
build cloud-based applications.
– IaaS: It dispenses services like storage, security tools, and networking for end users.

Several critical security attacks are designed and proposed against various cloud
deployment models which pose severe security risks to the adaptor. Attacks such as
wrapping, malware injection, and other attacks can be conducted against cloud
computing. One such security attack is the Denial of Service (DoS) attack. It is an
event or malicious behaviour that mitigates or prevents the cloud’s security to per-
form its expected functions and services. A distributed form of DoS attack is called a
Distributed Denial of Service (DDoS) attack which applies numerous network hosts
to inflict more effects on the dupe [2].
DDoS attacks can be classified into the subsequent categories
– Attacks on Bandwidth
– Physical Disruption.
– Limitation Exploitation
– Attacks on Connectivity
– Exhaustion of Resources
– Process disruption
– Data corruption
 Security and Privacy Preservation Model to Mitigate DDoS Attacks in Cloud | 43

Software-Defined Networks (SDNs) have emerged as a new communication para-

digm, released from direct integration into traditional networks and provides a sys-
tem and network flexibility through a sensible network controller. SDNs contain a
data plane, a control plane, and an application plane. The control plane contains
some controllers that use the logic control strategy and maintain the entire network
view as logic-centric. Administrators remove all network views from network ser-
vices and provide an easy-to-use interface for providers, researchers, or third parties
to facilitate these employees to customize transaction plans and see reasonable
network management. SDNs users do not have to worry about the technical details
of the device below, a simple system can detect the fast delivery of new applications
Advanced control features and systems enable SDNs to be exposed to known as
Distributed Denial-of-Service (DDoS) attacks. For example, the controller, which
plays a very important role in determining the performance of each component in
SDNs, is the main target of DDoS attacks. A compromised controller can lead to the
misconduct of all the switches under their control. Denial of service (DoS) attacks
can run out of system resources on a targeted computer, stop services, and leave its
normal users inaccessible. When hackers use two or more compromised computers
as a “doll machine” to launch a DDoS attack on a particular target, it is called a
DDoS attack. The IP address of the grinding machine and the composition and type
of attack package are not set, making it difficult to locate the attacker.
DDoS attacks have become a major threat to the Internet today and attacks
make online services unavailable to large victims on the road from multiple attack-
ers. With the number of businesses offering their services online growing exponen-
tially, a DDoS attack could lead to significant financial losses. A recent report re-
veals that DDoS is attacking a 22% data recovery account for 2015 [3-4].
DDoS attacks work in three steps, namely scanning, logging in, and launching
attacks. The unusual mobility of DDoS attacks often affects multiple paths and net-
work domains. The paper also deals majorly with KNN models which will be impro-
vised to detect attacks.
Cloud computing is one form of distributed system containing a set of virtual
machines that can be dynamically provisioned to meet the changing requirements
of resources by the customer. The cloud-customer relationship is governed by the
service level agreement (SLA).
Overhead of physical installation and maintenance of such a huge system issue
is relieved by the cloud thus reducing the overall costs and also increases efficiency.
Virtualization in cloud computing assumes Service-Oriented Architecture (SOA) that
enables the transformation of customer requirement problems into services thus
benefitting the idea of cloud computing. The cloud customer has to depend on the
Cloud Service Provider (CSP) for security and privacy of their information.
44 | Sirisha Potluri, Sachi Nandan Mohanty, T. Kundana, D. Abhinav, P. Sushrutha

Therefore, clearly there should be goals for the information security that should
be satisfied which are:
C – Confidentiality
I – Integrity
A – Availability
Confidentiality means making sure that the information that has to be kept secured
is secured and there shall be no access for unauthorized users to read the infor-
mation. Integrity means making sure that unauthorized users cannot modify or
destroy the information. Availability means making sure that authorized people are
not prevented from accessing the information.
Cloud architecture and its disparity from standardized traditional onsite system
makes the identification and separation of different aspects in cloud security an
arduous task.
This module is an attempt in exploring various security challenges and propos-
ing suitable solutions for cloud security.

2 Literature Analysis
In the DDoS attacks, the attacker dispatches an order to a system called as com-
mand and control server (C&C server) which will coordinate and trigger a botnet.
Generally, a botnet can be considered as a set of compromised hosts, which obscure
the attacker by providing a level of indirection. The C&C server orders the botnet to
launch a DDoS attack against the victim (server), and afterward, the bots will direct
the attack packets to the victim whose content depends on the type of attack. Thus,
the attacker host is separated from its victim by one or by tons of intermediate layers
of zombie hosts. A zombie or bot is a compromised computer under the control of an
attacker who controls many other machines which altogether form a botnet [5-7].

3 Classification of DDoS Attacks

Distributed denial-of-service attacks can be categorized based upon on its various
features and origin of an attack. DDoS attacks can be categorized as external DDoS
attacks and internal DDoS attacks [8-10].
 Security and Privacy Preservation Model to Mitigate DDoS Attacks in Cloud | 45

3.1 Various types of external and internal attacks

3.1.1 External attacks to internal attacks

In this case, the botnet which performs the attack comes from an exterior target
system. The attack can target a net entrée of the Cloud infrastructure or the servers.
If a specific client in a virtual machine becomes a victim of an attack, it’ll also affect
the opposite virtual machines present on the equivalent server of the Cloud.

3.1.2 Internal attacks to external attacks

In this case, the attack can tackle by taking possession of a virtual machine that is
running within the Cloud. This can be done with a Trojan horse. A Trojan horse is
malicious software or a code that can take control of the system. The choice of
which depends on the customer’s virtual machine to infect is important because if a
customer owns a large number of virtual machines, the Trojan horse can probably
spread all over the virtual machines, thus forming a botnet.

3.1.3 Internal attacks to internal attacks

In the Cloud infrastructure, an internal botnet is formed which can attack another
target inside the system such as a single virtual machine or a bunch of virtual ma-
chines. All the Cloud infrastructures may break down under these kinds of attacks.
Different kinds of attacks approach different types of attackers where each type of
attack scenario correlates to a particular attacker with a specific location and goals.

3.2 Classifications of DDoS attacks

3.2.1 Protocol vulnerability exploitation

These attacks take advantage of the known weaknesses which are present in the
protocol such as flaws in design or implementation of triggers, inappropriate activi-
ties, and can change the data that goes to and from a particular goal. Depending on
the design certain protocol steps might be created to check the potential for DoS
attacks.
46 | Sirisha Potluri, Sachi Nandan Mohanty, T. Kundana, D. Abhinav, P. Sushrutha

3.2.2 Malformed packets

A malformed packet attack are often launched against many protocols. For instance
malformed packet attack against IP protocol. In IP packet options attacks, mal-
formed packets can randomize the optional fields which are within an IP packet and
can set all the standard of service bits to at least one, which could cause more pro-
cessing within the victim for packet handling

3.2.3 Flooding attacks

Flooding attack is also called Bandwidth distributed Dos (BW-DDoS), the attacker
tries to flood the victim with unwanted traffic to prevent legitimate traffic. Based on
the type of protocol flooding attacks differs. Few of the strong attacking agents in-
clude privileged zombie which has control over its host whereas weak agents in-
clude programs that are downloaded automatically and run in sandboxes.

3.2.4 Amplification attacks

They are the devastating version of reflective DDoS attacks. They utilize the inherit-
ed nature of network protocols to increase the amount of traffic that is reflected in
the victim. As a result, the traffic that reaches the victim is amplified by the reflector
server. Amplification DDoS attacks are created by the amplification through flow
multiplication attacks or through payload magnification where a large number of
responses are reproduced or response messages bigger than the corresponding re-
quests are issued by the reflector

3.2.5 Reflective attacks

Another method that’s applied by the DDoS attackers is the reflective method. This
method helps the attacker to send traffic to the victim indirectly and helps the attacker
to stay undetected. During this method, all attack packets, which the attacker sends
contain the IP address of the victim within the source address field of IP packets.
Classification of DDoS attacks is given fig. 1.
 Security and Privacy Preservation Model to Mitigate DDoS Attacks in Cloud | 47

Reflective attacks

Flood packets

Protocol Attacks by
vulnerability exploited
exploitation
vulnerability

Amplification
attacks

Malformed
packets

Fig. 1: Classification of DDoS Attacks

3.3 Current popular DDoS attacks

– AgoBot: It is one among the foremost popular bots with an anti-virus vendor that
has over 600 versions. Its variants are Gaobot, Nortonbot, Phatbot, and Polybot
– SDBot: It’s over 1800 variants and comes with ping and UDP flooding tools.
– RBot: It has over 1600 variants and is written in C++ to focus on windows system.
– SpyBot: It’s written within the C programming language and also affects the win-
dows operating system.

3.4 Factors effecting severity of DDoS attacks

– Type of attack
– Type of protocol or event misused within the attack
– No: of attacking hosts
– Amount of resources at the victim’s site
– Type of cloud, mostly public cloud
48 | Sirisha Potluri, Sachi Nandan Mohanty, T. Kundana, D. Abhinav, P. Sushrutha

4 DDoS Mitigation Techniques

DDoS mitigation techniques help us in effectively combating the attacks. Few of the
mitigation techniques are as follows:

4.1 Reducing attack surface

Reducing the services which will be attacked limits the choices for attackers. Using
Firewalls and Network Access Control Lists will allow only necessary traffic, to nec-
essary ports from essential hosts only. Considering the case of web servers, they
permit traffic from anywhere to port 80 of your webserver. And in such cases, one
should further take other protective measures. Even for the websites which are ac-
cessible across the web, one can reduce the area by restricting traffic to countries
where users are located.

4.2 CDN’s
CDN’S distributes the content and boosts the performance by minimizing the space
between resources and between the end-users. It stores the cached version of the
content in multiple locations and this eventually mitigates DDoS attacks by avoid-
ing one point of failure, when the attacker is trying to pivot on a single target.

4.3 Black hole routing

As the name propounds, black hole routing, with none of the filtering routes both
legitimate and malicious traffic route to a null route or black hole where the traffic
will be dropped from the network. Depending on the pattern, if one can identify the
attacker, then the packets are often filtered and can be routed back to the black
hole.

5 Stages of DDoS Mitigation

5.1 Detection
Initially there is a need to detect if an attack is present or not. At this phase, there is
a need to identify the legitimate traffic and malicious software. In the event of miti-
gating DDoS one shouldn’t drop potential customer traffic which would be disas-
trous.
 Security and Privacy Preservation Model to Mitigate DDoS Attacks in Cloud | 49

5.2 Response
Once the attack is detected now there is a necessity to seek out how to respond to
those attacks.

5.3 Routing
By intelligent routing, one can break the remaining traffic into manageable chunks
which will be handled by cluster resources to which it’s being routed.

5.4 Adaption
This is the of the foremost important stage of DDoS mitigation where one has got to
search for patterns of DDoS attacks and check if this analysis can further strengthen
the mitigation techniques.

6 Privacy in DDoS Attacks

ANALYSIS OF DATASET LLDOS is as follows:

Stage Flow Number 172.16.112.* 172.16.113.* 172.16.114.* 172.16.115. *

Victim Attacker Combine Scanning 424 328 32 296 0 1081 1081
Intrusion 128 0 97 0 0 332 335
Attacking 2 0 0 285 107465 199 107667

The KNN features for ease of use are very useful for us when we want to embed other
special functions in the separator to protect the privacy of test data. On the other
hand, KNN, as a partitioning algorithm, works well with precision, and KNN is not
sensitive to vendors who can maintain high accuracy when there are certain sounds
in the database.
The European vector n dimensional formula is:

𝑑 ∑ 𝑋 𝑋 ∗2 (1)

KNN as high precision is widely used in a lot of different areas.

50 | Sirisha Potluri, Sachi Nandan Mohanty, T. Kundana, D. Abhinav, P. Sushrutha

7 Summary of Our Contributions Using KNN

– We recommend Predis, a DDoS privacy protection program for SDNs privacy,
which looks at both the detection of DDoS attacks and privacy protection in
multi-party partnerships. Predis uses SDN features and an improved KNN algo-
rithm to detect DDoS attacks accurately during real-time and integrates digital
cryptography and perturbation encryption to provide each participant’s privacy
and confidentiality.
– We attest to Predis’s security in the asymptotic form of computer security in
modern cryptography. By strict security analysis, we confirm that the traffic da-
ta provided by each participant is inseparable from its potential adversary.
– We do extensive research using the most authoritative data sets to show that
Predis is timely and accurate. We show that our system can not only determine
if traffic is normal, but can also detect unusual traffic at the beginning of a
DDoS attack. The results show that Predis is more accurate than the existing
adoption schemes, which currently protect participant’s privacy. A detection at-
tempt restricted within one domain would be unable to spot the attacks at their
primary stages. Thus, the involvement of multiple domains in attack detection
will help to realize more accurate and timely detections.

8 Summary of DDoS Attacks Detection Methods

There are no. studies on DDoS attack detection because of the prevalence and sever-
ity of DDoS attacks. So we briefly summarize related work from 2 perspectives, i.e.-
DDoS attacks detection in conventional networks and DDoS attacks detection in
SDNs. Detection approaches of DDoS attacks have been researched exceedingly in
conventional network models [11-12].

8.1 Conventional network entropy

– Support Vector Machine
– Naive Bayesian
– Neural Network
– Cluster Analysis
– Artificial Neural Network
– K-Nearest Neighbours
 Security and Privacy Preservation Model to Mitigate DDoS Attacks in Cloud | 51

8.2 SDNs
– Self- organizing map B
– Support Vector Machine
– Entropy variation of the destination IP address
– Deep Learning
– Bayesian Networks

The attackers of DDoS can simultaneously control several computers and make an
attack architecture that contains control puppets/dummies and attack pup-
pets/dummies in the computing environment. The traditional attack architecture is
analogous to the dumbbell shape structure. Wherein an intermediate network is
mainly responsible for data forwarding, security events and control functions are
done by the management, while the network itself can’t detect network attacks and
deal with them.

9 Privacy Preserving in Cross-Domain Detection

A SDNs domain in Predis refers to the controlled domain which is under SDNs archi-
tecture, which is in the network domain along with deployment of the SDNs tech-
niques and can independently get controlled by operators. The SDNs domains con-
duct a centralized way of controlling data forwarding. The multiple SDNs domains
explain in our paper collaborate and these SDNs domains may or may not be adja-
cent to each other physically or in geographical location. Control plane of this do-
main sends the flow table to a selected location on the computing server. The com-
puting server provides the DDOS detection service and returns detected results to
controllers.
The traditional network domain for traffic forwarding is a distributed control
where it cannot achieve centralized control. Privacy-preserving cross-domain at-
tacks detection are often seen as a Secure Multiparty Computation (SMC) problem,
which is that the matter of the way to calculate a function safely when no credible
third party is present there [13-14].
KNN algorithm occurs in embedding encryption steps. After providing training
details, KNN can separate test samples by selecting a distance measurement formu-
la outside the training phase (Euclidean grade selected in Predis).
52 | Sirisha Potluri, Sachi Nandan Mohanty, T. Kundana, D. Abhinav, P. Sushrutha

10 Models
10.1 System model/approaches and threat model
We first go through the overview of the system model/approaches and the roles
included in Predis. Then we present the threat model, followed by security.
It has three sorts of roles:
– Server for Computation (CS)
– Server for Detection (DS)
– Software Defined Networks SDNs domain 3

Domain D n is the nth domain that participates in the attack detection and provides
the data to CS and DS. Which in turn, provides computing & encryption services for
domain Dn. Each domain sends traffic information to CS for the calculation purpose
and receives the detection results from DS. CS provides computational service and
sends the intermediate results to DS. Where this later provides detection service
based on intermediate results and replies to the detection results of each domain.
Thus, CS and DS perform computation in collaboration with each other. Predis pro-
vides accurate DDoS attacks detection service for domains wherein each domain is
not willing to share privacy traffic information.
Here, we provides a formal definition of privacy. The knowledge of the flow ta-
ble is provided by domains that participate within the detection is called privacy.
Importantly, privacy includes IP Source, IP Destination, Port Source, Port Destina-
tion, Length, and Flow Packets. The basic operations includes these three roles
mentioned earlier as functions with input and output. Each function is meant to run
on continuous inputs in real-time data partitioned into a particular interval and
Predis features a set of n input peers with it. Sample code of DDoS is given as below.

Listing 1: Performing DDoS attack for service inaccessibility

import socket
import sys as s1
import os
print("DDoS Attacking: "+s1.argv[1]+" Initiating...")
print("Inject Code "+s1.argv[2])
def attackFunc():
s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.connect((s1.argv[1], 80))
print(">>> GET Message /" + s1.argv[2] + " HTTP/1.1\n")
s.send("GET Message /" + s1.argv[2] + " HTTP/1.1\n")
s.send("From Host Initiate: " + s1.argv[1] + "\n");
 Security and Privacy Preservation Model to Mitigate DDoS Attacks in Cloud | 53

s.close()
for var in range(1, 10000):
attackFunc()

11 Security in Different Cloud Computing Service

Models
Three service models of cloud that provide services at different layer are Software as
a service, Platform as a service and Infrastructure as a service respectively.
SaaS – Customer has minimum command on security as the backend infrastruc-
ture and also the platform of execution is present outside the range of the customer.
PaaS – In PaaS, the customers have a certain amount of command on the deployed
applications and remotely on configuration settings for the application environ-
ment. This framework gives more control over security compared to SaaS but less
control than what IaaS gives. IaaS – In IaaS, the customer has much more control
over security than the two other models as IaaS delivers virtual computing resources
as networking, storage devices, and hardware.

12 Security in Different Cloud Deployment Models

The four deployment models of cloud out of which the customer selects based on
suitability, requirement, and purpose of the user cloud are Public, Private, Commu-
nity Cloud, and Hybrid clouds.
Public Cloud – Organizations have to compromise certain aspects of security in
this deployment model as the cloud lies at the service provider end. Customers are
unaware of the geographical location of their data, type of storage used by CSP
hence making this model a bit less secure.
Private Cloud – It is owned by a single organization or by an on or off-premise
located the third party. This deployment model solves a few aspects of security is-
sues faced in the public cloud to some extent but has its difficulties like provision-
ing, storage management, and capacity watching.
Community Cloud – It is owned by a community of organizations that have a
common interest. The drawback of this model is that there are several unanswered
questions in terms of data proliferating across multiple domains and organizations
and service outrages.
Hybrid Cloud - As a hybrid cloud is a combination of any of the two deployment
models, it covers the benefit of scaling, cost, and also the security. Data privacy and
integrity form a threat in the hybrid model.
54 | Sirisha Potluri, Sachi Nandan Mohanty, T. Kundana, D. Abhinav, P. Sushrutha

13 Vulnerabilities in Cloud
Virtualization is the backbone to cloud and it comes in three levels. First at the op-
erating system level, second at application level and finally at hypervisor level.
In the operating system level, multiple guest operating systems run on the host
operating system making guest operating systems vulnerable as the control is at the
host operating system. If the host OS is compromised, then so is the guest OS.
The application level is located just above the host OS layer. Hence the virtual
machines which run guest OS along with its subsequent applications also suffer
from the same vulnerability which is present at the operating system level.
At the hypervisor level, as the virtual monitor (VMM) or the hypervisor which is
a software layer runs on the host operating system, every other virtual machine
which runs on guest operating systems are controlled by the hypervisor at host OS
making it a vulnerable case because when hypervisor at host OS is compromised,
then so are virtual machines at guest OS. A small flaw at the hypervisor level creates
risk for the entire infrastructure of the cloud; also the breaches in hypervisor create
cross virtual machine attacks.
Since the cloud services are acquired by its users through the application inter-
faces (APIs) provided by cloud service providers, we can say that the security of
these APIs directly affects the security of the cloud. Loopholes in TCP/IP protocol
explicitly cause vulnerabilities typically to public clouds

13.1 Threats
As the computing infrastructure of the cloud is present off-premises, the organiza-
tion or the users who have opted for the cloud services must assess the risk involved
when there is a loss of control of these services.
In the beginning stages of users acquiring the cloud services, there is a chance
that with insufficient knowledge, they can set the whole infrastructure of cloud at
risk by improper usage of storage capabilities, processing, and by the deployment of
dubious applications. This kind of threat is highly relatable to PaaS and IaaS.
Also by knowing the entire or part of the system, an attacker knows how to go
through security mechanisms. The attacker also can have privileges to bypass fire-
walls and also for intrusion detection systems. Virtualization, sharing of resources,
and multi-tenancy are the concepts that also help the attacker in exploiting the
loopholes and give the ability to manipulate the service models.
The data in cloud services traverse across the continents where there is a chance
for issues arising in terms of jurisdiction as different geographic locations have dif-
ferent obligations in laws than the service provider.
 Security and Privacy Preservation Model to Mitigate DDoS Attacks in Cloud | 55

Deceptive mechanisms and weak passwords may lead to the user being the suf-
ferer of identity theft.
An attacker can launch denial of service attacks by flooding the users with re-
quests, can attack virtualization using hypervisor root kits, can do phishing, and
can do meta-data spoofing with the above-mentioned vulnerabilities and threats.

14 Security Issues
14.1 Security issues layer-wise
Cloud can be viewed in distinct layers such as a physical, Virtualization, Service
provider and User layer. There are certain security issues associated with each layer.

Tab. 1: Security issues in cloud computing architecture layer-wise

Layer Description Security issues

Physical layer Provides computing, storage, 1. Database intrusions

and networking resources. 2. Data storage issues
3. Confidentiality, integrity,
availability.
4. Network vulnerabilities and
attacks
Virtualization layer The hypervisor is present at 1. Access control issues
this layer along with the guest 2. Regulations compliance
operating systems and virtual 3. Hypervisor and virtual
machines machines exposure
4. Isolation between virtual
machines
Service provider layer Offers management services 1. Trust management
such as scheduling, load bal- 2. Transient data security
ancing, accounting policy issues
management, resources provi- 3. Authentication and authori-
sioning, and monitoring. zation issues
4. Audit, regulations compli-
ance
5. Policy enforcing
User layer Have applications, application 1. Authentication, access
interfaces, and tools. control
2. Exposure of browser and
API’s
3. Vulnerabilities in applica-
tions
56 | Sirisha Potluri, Sachi Nandan Mohanty, T. Kundana, D. Abhinav, P. Sushrutha

15 Proposed Solutions in Literature

Various state of the art methods in literature are listed as follows.

15.1 Data confidentiality schemes

Tab. 2: Data confidentiality schemes

Name of the scheme Proposed by Idea of the scheme Algorithms used

Searchable Encryption I-En Liao & Jyun-yao Secret sharing and 1. Non-numeric file
Hung searchable encryption search
(2012) 2. Data Encryption
Onion Encryption Curino et al. Adaptable security 1. Query Execution
(2011) providing different 2. Data Encryption
layers of encryption in
order to maintain the
confidentiality of the
user data
Fully Homomorphic Tebba et al. Allowing the clients in 1. Encryption
Encryption (2012) encrypting their data 2. Key - Generation
before storing at cloud 3. Evaluation
service provider there-
by ensuring the data
confidentiality
Encryption for secure Yu et al. To Protect user data by 1. Key - Generation
scalable fine grained (2010) using cryptographic 2. Setup
access control based techniques and also to 3. Encryption
on attributes induce overheads of 4. Proxy Re
key distribution and 5. Decryption
management on the
client data.

15.2 Cloud virtualization confidentiality schemes

Tab. 3: Cloud virtualization confidentiality schemes

Name Of The Scheme Proposed By Idea of the Scheme Algorithms Used

PALM Zhang et al. Prototype that ensures 1. Metadata Migration

(2008) security of protected Protection
user data and backup 2. Migration data
data while and after Protection
 Security and Privacy Preservation Model to Mitigate DDoS Attacks in Cloud | 57

Name Of The Scheme Proposed By Idea of the Scheme Algorithms Used

VM live migration
Trusted cloud compu- Santos, Gumadi and Targets at preserving 1. VM Launch
ting platform Rodrigues confidential execution 2. Node Registration
(TCCP) (2009) of VM’s
Self-service cloud Ganapathy V To solve the issues 1. Create_MTSD
computing scheme (2015) related to uninterrupt- 2. Create_UDom()
(SSC) ed cloud service pro- 3. Cre-
vider access on the ate_Userdomain
client CPU contents, 4. Bootstrapping_SSL
memory and registers. 5. Grant_Privilege
TVDC IBM Concept in which each 1. Authorization of
(2008) VM is restricted from VMM
accessing any other 2. Resource Access
VM thus ensuring 3. Infrastructure
protection from un- Integrity and
wanted data leakage Network Isolation
4. Inter VM
communication

15.3 Cloud data integrity schemes

Tab. 4: Cloud data integrity schemes

Name of the scheme Proposed by Idea of the scheme Algorithms used

Public verifiability and Wang et al. To provide cloud data 1. KeyGen

data dynamics scheme (2009) storage integrity 2. SigGen
3. GenProof
4. VerifyProof
5. ExecUpdate
6. VerifyUpdate
MHT Niaz M.S, Saake Gin Merkle’s scheme for 1. Multi-Join
(2015) maintaining user data 2. Single-Join
integrity without the 3. Zero-Join
overhead of maintain- 4. Range Condition
ing a table at owner
side of data.
Dynamic provable data Erway et al. Supports data sponta- 1. Prepare Update
possession (2009) neity with cloud data 2. Verify Update
integrity 3. Perform Update
4. Challenge
5. Proof
6. Verify
58 | Sirisha Potluri, Sachi Nandan Mohanty, T. Kundana, D. Abhinav, P. Sushrutha

Name of the scheme Proposed by Idea of the scheme Algorithms used

Privacy preserving Wang et al. To assure data integri- 1. KeyGen

public auditing scheme (2015) ty of cloud storage and 2. SigGen
providing an optimal 3. GenProof
data verification meth- 4. VerifyProof
odology to prevent
integrity breaches

15.4 Integrity schemes in cloud virtualization

Tab. 5: Integrity schemes in cloud virtualization

Name of the scheme Hypervisor used Algorithms used

ACPS KVM 1. Checking Activity

2. Logging of Activity
3. Checksum calculation
4. Alert Generation
5. Generating Security Response
SSC XEN 1. Create_UDom
MIRAGE VMware 1. Access Control
2. Image Transformation
3. Provenance Tracking
4. Image Maintenance
PALM Xen 1. Data Protection Migration
2. Metadata Migration Protection

16 Onion Encryption (OE)

This is a data confidentiality scheme. It is an outlook to an adaptable security with
distinct layers of Encryption (similar to layers of an onion) so that the SQL queries
can be executed on the encoded data which includes aggregates, operations of or-
dering. joins and the processing of query is done at the cloud service provider’s end
simultaneously maintaining user data secrecy as decryption will happen at the cli-
ent end.
The primary focus is to maintain distinct levels of encryption for every column
and decoding every column as per the requested query.
This scheme uses two algorithms
– Data Encryption Algorithm
– Query Execution Algorithm
 Security and Privacy Preservation Model to Mitigate DDoS Attacks in Cloud | 59

These algorithms take the use of keys such as RND, DET, OPE, HOM.

1. Randomized Encryption Key (RND): It provides similarity under an adaptive chosen plain text
attack.
2. Order preserving Encryption Key (OPE): For the queries which involve choices based on compari-
son, this key provides secured execution.
3. Homomorphic Encryption Key (HOM): For the queries which involve server side aggregates calcu-
lation, this key is used for execution.
4. Deterministic Encryption Key (DET): Queries which involve a selection on equality for a specific
value, this key is used for execution.

Time complexity: Time complexity can be calculated as

If T1 = time spent in redrafting queries
T2 = Time required in encoding and decoding payloads
Time complexity is the order of T1 * T2 i.e., O (T1*T2)
It has been discovered that onion encryption scheme brings a comprehensive drop
in throughput by 22.5%.

17 Conclusion
Cloud computing comes with known vulnerabilities because of the combination of
technologies like network links, web servers, virtual machines, etc., since cloud
computing has an advantage of scalability and elasticity it offers adequate re-
sistance to attacks. Among those, the DoS and DDoS are easy to escalate and are
more destructive. Security should be considered as shared responsibilities of both
cloud providers and users. Traditional computing has been completely transformed
to a cost effective, optimized computing by cloud computing. Yet there are some
challenges and issues in regard with secured information sharing in cloud. Existing
security measures needed to be enhanced and also new security mechanisms are
needed in order to eradicate security issues in cloud environment.
In this paper few of the mitigation techniques, solutions, and schemas are pre-
sented through which the security and privacy of the cloud can be preserved. Few of
the solutions could not perfectly mitigate all the possible attacks, but other solu-
tions are efficient. An SDN-based cross-domain attacks detection scheme was pre-
sented that provides privacy protection. Also, we came across cross-domain privacy
protection problems and also the DDoS attacks detection based on SDNs, in which
we combined geometric transformation and data encryption methods in the inten-
tion to protect privacies, broke down the detection process into two steps, disturb-
ance, and detection, which involved two of the servers that work together to com-
plete the detection process by using the concept of cloud virtualization. An
60 | Sirisha Potluri, Sachi Nandan Mohanty, T. Kundana, D. Abhinav, P. Sushrutha

enhanced KNN algorithm for low time consumption and high accuracy was repre-
sented. Extensively experiment of different authors results showed that Predis is
proficient in detecting cross-domain anomalies while preserving privacy with the
best time complexity and high accuracy. The overall paper can be used to further
reduce the time consumption of Predis in attack detection in the future.

18 References
[1] Saurabh S, Young-Sik J, Hyuk J, A survey on cloud computing security: issues, threats, and
solutions, journal of network and computer applications, 2016, Volume 75, Pages 200-222,
ISSN 1084-8045, https://doi.org/10.1016/j.jnca.2016.09.002
[2] Shankarwar M, Pawar A, Security and privacy in cloud computing: a survey, Satapathy S,
Biswal B, Udgata S, Mandal J, Proceedings of the 3rd International Conference on Frontiers of
Intelligent Computing: Theory and Applications (FICTA) 2014, Advances in Intelligent Systems
and Computing, 2014, vol 328, Springer, Cham. https://doi.org/10.1007/978-3-319-12012-6_1
[3] Andrew C, Mohammad H, Omar A, Defence for distributed denial of service attacks in cloud
computing, Procedia computer science, 2015, Volume 73, Pages 490-497, ISSN 1877-0509,
https://doi.org/10.1016/j.procs.2015.12.037
[4] Mohammad M, Marzie J, A survey and taxonomy of DoS attacks in cloud computing, Sec. and
Commun. Netw, 2016, 9, 16, 3724–3751, DOI:https://doi.org/10.1002/sec.1539
[5] Potluri S, Mangla M, Satpathy S, Mohanty S N, Detection and prevention mechanisms for DDoS
attack in cloud computing environment, 2020 11th International Conference on Computing,
Communication and Networking Technologies (ICCCNT), Kharagpur, India, 2020, pp. 1-6, doi:
10.1109/ICCCNT49239.2020.9225396
[6] Priyanka K, Munesh T, Virendra Y, Vikash S, Detection techniques of DDoS attacks: a survey,
2017 4th IEEE Uttar Pradesh Section International Conference on Electrical, Computer and Elec-
tronics (UPCON), Mathura, India, 2017, pp. 675-679, doi: 10.1109/UPCON.2017.8251130
[7] Zhang P, Huang X, Sun X, Wang H, Ma Y, Privacy-preserving anomaly detection across multi-
domain networks, 2012 9th International Conference on Fuzzy Systems and Knowledge Dis-
covery, Chongqing, China, 2012, pp. 1066-1070, doi: 10.1109/FSKD.2012.6234272
[8] Jelena M, Peter R, A taxonomy of DDoS attack and DDoS defense mechanisms, SIGCOMM
Comput. Commun, 2004, Rev. 34, 2, 39–53, DOI:https://doi.org/10.1145/997150.997156
[9] Soule A, Ringberg H, Silveira F, Rexford J, Diot C, Detectability of traffic anomalies in two adja-
cent networks, Uhlig S, Papagiannaki K, Bonaventure O, Passive and Active Network Measure-
ment, PAM 2007, Lecture Notes in Computer Science, 2007, vol 4427, Springer, Berlin, Heidel-
berg, https://doi.org/10.1007/978-3-540-71617-4_3
[10] Yaar A, Perrig A, Song D, Pi: A path identification mechanism to defend against DDoS attacks,
2003 Symposium on Security and Privacy, 2003, Berkeley, CA, USA, pp. 93-107, doi:
10.1109/SECPRI.2003.1199330
[11] Shui Y, Yonghong T, Song G, Dapeng W, Can we beat DDoS attacks in clouds?, IEEE Transac-
tions on Parallel and Distributed Systems, 2014, vol. 25, no. 9, pp. 2245-2254, doi:
10.1109/TPDS.2013.181
[12] Keromytis A, Misra V, Rubenstein D, SOS: An architecture for mitigating DDoS attacks, IEEE
Journal on Selected Areas in Communications, 2004, vol. 22, no. 1, pp. 176-188, doi:
10.1109/JSAC.2003.818807
 Security and Privacy Preservation Model to Mitigate DDoS Attacks in Cloud | 61

[13] Mousavi S, St-Hilaire M, Early detection of DDoS attacks against SDN controllers, 2015 Interna-
tional Conference on Computing, Networking and Communications (ICNC), Garden Grove, CA,
USA, 2015, pp. 77-81, doi: 10.1109/ICCNC.2015.7069319
[14] Saied A, Richard O, Tomasz R, Detection of known and unknown DDoS attacks using artificial
neural networks, Neuro computing, 2016, Volume 172, Pages 385-393, ISSN 0925-2312,
https://doi.org/10.1016/j.neucom.2015.04.101-393
Sirisha Potluri, S. Sunaina, P. Neha, Ch. Govind, J. Raghavender,
V. MNSSVKR Gupta
A Secure Cloud Infrastructure towards Smart
Healthcare: IoT Based Health Monitoring
Abstract: The magnitude of big data is estimated to grow in the field of healthcare.
Big Data when focused on the sector of health care can be best described as struc-
tured, unstructured and semi structured way of representing a patient’s personal
data such as his/her medical conditions, prescriptions, etc. Handling and pro-
cessing such huge chunks of data can be a daunting task. To avoid scenarios of this
type, many of the healthcare sectors have adapted cloud computing as the solution.
Amidst the pandemic such as the novel covid-19, remote working has become the
new normal. Healthcare sectors are gradually adapting to the idea of remote treat-
ment also known as telemedicine. A patient’s medical condition has to be coordi-
nated with a team of specialists, nutritionists, nurses, etc. By using cloud computing
one can completely avoid a physical interaction, it’s very useful in current situations
like covid-19. Hence cloud computing is chosen by many health sectors to receive
and store huge chunks of patients data and manage their electronic records. This
huge chunk of electronic records helps in the process of mining the data, identifying
the patterns and developments in the big data healthcare sector. Main things that
are to be taken care of : Privacy related issues and Security related concerns, Dy-
namically acceptable storage, Regulatory issues , Electronic documents containing
patients data, handwritten medical editions ,various images such as X-ray, MRI
scan, radiology images, etc. can be retrieved at any time and from any place. This

||
Sirisha Potluri, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana
501203, India, [email protected]
Sai Lalitha Sunaina, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana
501203, India, [email protected]
Neha Pavuluri, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI Foundation
for Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana 501203, India,
[email protected]
Chennu Sai Sri Govind, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana
501203, India, [email protected]
Raghavender Rao Jakileti, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana
501203, India, [email protected]
V. MNSSVKR Gupta, Department of Computer Science Engineering, SRKR Engineering College,
Chinnamiram, Bhimavaram, Andhra Pradesh 53420, India, [email protected]

https://doi.org/10.1515/9783110732573-004
64 | Sirisha Potluri, S. Sunaina, P. Neha, Ch. Govind, J. Raghavender, V. MNSSVKR Gupta

can be an advantage and a disadvantage at the same time. The main challenges
involved are failures of the network, security related concerns, and privacy related
issues of patient information that is being ill-treated by users/customers, hackers,
network downtime, malware, and etc.

Keywords: Cloud Computing, Virtualization, Data Security, Cryptography, Denial of

Service, Health Care, Big Data, Blockchain

1 Introduction
Cloud computing is a distributed computing model, helps in delivery of computer
system services including computing power and data storage. Cloud model uses
dynamically provisioning of resources to meet the varying and dynamic resource
requirements of a customer by the concept of distributed systems in which various
virtual machines are grouped together. Virtualization is the core and has an im-
portant role in cloud computing. Virtualization technology involves creation of a
virtual version of a server, a storage device, a desktop, an operating system. Pro-
gramming level virtualization helps in PaaS offerings whereas IaaS is based on the
idea of hardware virtualization [1].
In the last decade, information technology has played a major role in the field of
healthcare. Many of the healthcare sectors have adapted cloud computing as the
solution. Hospitals and healthcare providers are exempted from the need to pur-
chase the hardware and servers. You can have huge cost savings by only purchasing
the resources you actually use with the help of cloud. Patients can have control over
their own health by using Cloud computing which democratizes data. By storing
Patient records and medical images on the cloud, data can be easily retrieved. With
help of Cloud computing healthcare providers can give prescriptions and treatment
protocols by easily gaining access to the patient data. Irrespective of geographical
locations Cloud computing also helps the specialists to review patient cases and
give their opinions. A number of healthcare-related utilities such as telemedicine,
pre- hospitalization, post-hospitalization etc. can be improved by applying cloud
computing models in healthcare [2].
The primary concern in the healthcare practices and observations for online ac-
cess of the EHR record -Electronic Health Record is Security and privacy. In order to
protect the patient’s information the sensitive records of Healthcare data should
not be made available to unauthorized people. However cyber gaps in cloud compu-
ting posture an adversarial impact and influence on the security and privacy of
patients’ electronic health records. The online access of patient records and transac-
tions related to diagnosis have many benefits for patients as well as healthcare or-
ganizations and professionals but it also raises serious security and privacy issues
related to private data of patients. Due to risk associated with EHR record -Electronic
 A Secure Cloud Infrastructure towards Smart Healthcare: IoT Based Health Monitoring | 65

Health Record it is important to guarantee the privacy of patients. Hacking incidents

on EHR record -Electronic Health Record systems may lead to modifying the data of
patients.
The advent of big data has caused multiple obstacles in various research fields.
The term “big data” refers to the vast and huge volume of structured and unstruc-
tured data. It consists of patient’s data and helps them to select the prime possibility
and also to support the medicinal treatment plan. However, they have made hard to
manipulate and process the data as the healthcare data became much larger. Con-
trastingly, the processing of big data on distributed sources is done by MapReduce.
One of the main reason for faults in healthcare domains is the incomplete and im-
perfect access to patient-related data and information and unsuccessful communi-
cation among the stakeholders involved in the process [3].

2 Cloud Deployment Models in Healthcare

Cloud deployment models are represented in fig. 1 as shown below.

2.1 Public cloud

This type of cloud model is available for anyone who wants to use or purchase them.
It is based on a shared cost model or pay per use policy.

2.2 Private cloud

This model is mainly used by stand-alone organizations. Private clouds are best for
organizations that require high-security and high management demands.

2.3 Community cloud

Community cloud is a shared model that is limited to only a set of organizations or
employees. This type of deployment model of cloud computing is managed and
hosted internally or by a third-party vendor.

2.4 Hybrid cloud

Hybrid cloud combines two or more models such as public cloud, private cloud or
community cloud. Based on business requirements Hybrid cloud architecture allows
66 | Sirisha Potluri, S. Sunaina, P. Neha, Ch. Govind, J. Raghavender, V. MNSSVKR Gupta

an enterprise to move data and applications between private and public environ-
ments.

Cloud Deployment Models

Public Cloud (Used by

general public)

Private Cloud (Used

by single
organization)

Community Cloud
(Used by
organizations with
shared interests)

Hybrid Cloud
(Combination of
clouds)

Fig. 1: Cloud deployment models for healthcare

3 Cloud Service Models in Healthcare

Cloud service models are represented in fig. 2.

3.1 Software as a service (SaaS)

SaaS describes a cloud service where consumers are able to access software applica-
tions running on a cloud infrastructure, over the internet. SaaS solutions do not
 A Secure Cloud Infrastructure towards Smart Healthcare: IoT Based Health Monitoring | 67

require the client to maintain anything, including the storage of data and the net-
work and are managed by the vendor.

3.2 Platform as a service (PaaS)

Platform as a Service (PaaS) is an integrated and abstracted cloud-based model that
supports the development, and management of applications. The whole storage and
management of data is left to the vendor.

3.3 Infrastructure as a service (IaaS)

IaaS is the virtual delivery of computing resources in the form of hardware, network-
ing, and storage facilities. Based on the amount of storage and the amount of pro-
cessing power you use over a time period you will be charged according to that.

4 Factors to be Considered for the Evaluation of the

Cloud System Security
4.1 Confidentiality
Confidentiality refers to protection of data from unauthorized users. In a Cloud Sys-
tem such users might want to get unauthorized access to the data of some other
individual. Patient’s information should be kept confidential in healthcare settings.
Patient’s reputation, opportunities, and human dignity may be under threat with
inappropriate disclosure of that information.

4.2 Integrity
Integrity guarantees that the security property of an asset has not been modified by
some unauthorized third-party personnel. It maintains the accuracy, reliability and
consistency of data throughout its entire ‘life-cycle.’ Integrity ensures an asset’s
accuracy and correctness with respect to its owner. In healthcare, the accuracy of a
patient’s personal details, health summary, clinical notes, test results etc. can be
maintained by integrity.
68 | Sirisha Potluri, S. Sunaina, P. Neha, Ch. Govind, J. Raghavender, V. MNSSVKR Gupta

4.3 Availability
One of the most important aspects of security that needs to be maintained is Availa-
bility. A service-level agreement is a bond between the customer and the service
provider about services to deliver in terms of availability and response to demand.
Earlier, the SLA were negotiated between the service consumer and a client. But
today there are different levels of SLA like customer based SLA, Service based SLA,
Multi-Level SLA.

Cloud Service Models

Software as a service
(SaaS)- Ex: Communication,
email and games

Platform as a Service
(PaaS)- Ex: Web, application
development

Infrastructure as a service
(IaaS)- Ex: Server, storage,
network

Fig. 2: Cloud service models for healthcare

Three important factors to be considered for the evaluation of the cloud system se-
curity are represented in fig. 3.
 A Secure Cloud Infrastructure towards Smart Healthcare: IoT Based Health Monitoring | 69

Integrity

Confidentiality Availability

Factors to be
considered for
cloud system
security evaluation

Fig. 3: Factors to be considered for the evaluation of the cloud system security

5 Review of Literature
Various security and privacy preservation methods and approaches in health care
are discussed in table 1 as shown below.

Tab. 1: Security and privacy preservation approaches in cloud computing environment

S.No Author Title Proposed Model

1 Jan de Muijnck-Hughes Predicate Based Encryption Predicate Based Encryption(PBE)

(2011) focuses on both the Platform as
a service and the Software as a
service implementation
2 Venkata Sravan et al. Security Techniques for Protect- Understanding security risks and
(2011) ing Data in Cloud the required security strategies
used in Cloud computing to
mitigate them
3 Ali Asghari Karahroudy Security Analysis and Frame- Partially Distributed File System
(2011) work of Cloud with Parity (PDFSP)
Computing
70 | Sirisha Potluri, S. Sunaina, P. Neha, Ch. Govind, J. Raghavender, V. MNSSVKR Gupta

S.No Author Title Proposed Model

4 Nabil Giweli Enhancing Data Privacy and Data Centric Security approach
(2013) Access
Anonymity in Cloud Computing
5 Miao Zhou Enhanced Privacy Cloud key- Innovative tree-based key man-
(2013) word search and Remote public agement scheme
integrity audit for private data
6 Sudhansu Ranjan Enhancing Data Security in The RSA algorithm is used for
Lenka et al. Cloud encrypted correspondence and
(2014) Computing using RSA Encryp- file encryption and decryption
tion and MD5 Algorithm while the digital signature of the
MD5 algorithm is used.
7 Aastha Mishra Security Advanced Secret Sharing Key
(2014) and privacy of user’s data Management Scheme
8 Nesrine Kaaniche Cloud Data ID-Based
(2014) Storage Security based on Cryptography (IBC) and CloudSec
Cryptographic
Mechanisms
9 Afnan Ullah Khan Data Confidentiality and Risk Access Control and Data Confi-
(2014) Management in Cloud Compu- dentiality(ACDC)
ting
10 Sarojini et al. (2016) Trusted and Reputed Services Enhanced Mutual Trusted Access
Using Enhanced Mutual Trusted Control
and Reputed Access Control Algorithm (EMTACA)
Algorithm in Cloud

6 Recommendations for Smart Healthcare Model

Healthcare has one way of communicating with a patient i.e. through the internet via
mobile devices and Laptop Considering the factors such as economic situations, loca-
tion and affordability mobile devices are in demand. Since our main focus is to pro-
vide a secured and reliable healthcare service to a patient with the support of cloud
computing services and facilities, we take the help of Paas, SaaS and BDaaS meaning
Platform as a service Software as a service Big Data as a service respectively.
Analytical mobile applications helps us to analyze and filter the collected data
beforehand so this is an added advantage compared to the normal mobile applica-
tions. These are built and hosted using cloud computing technologies. Mobile ap-
plications are accessed through web servers that run on mobile browsers without
the considerations of mobile specific operating system, capacity and memory.
Cloud computing technologies includes delivery of various types of facilities as
shown in fig. 4.
 A Secure Cloud Infrastructure towards Smart Healthcare: IoT Based Health Monitoring | 71

– SaaS (Software as a service): The provider platform consists of a residing soft-

ware and the consumer can access it using a web browser or an API. It is a pay
as you go model. Salesforce and office 365 are the examples
– PaaS (Platform as a service): An application is developed by the consumer and
is hosted on a virtual server since the consumer has some control over data and
applications it pushes to develop, test and deploy applications very fast.
– IaaS (Infrastructure as a service): IaaS is the virtual provisioning of the compu-
ting resources such as hardware, networking, and storage facilities.
– BDaaS (Big Data as a service): It delivers statistical analysis tools by an outside
provider which helps the organizations to gain insights from a large information
set to gain an advantage.
– NaaS (Network as a service): Network as a service defines services and facilities
for network transport connectivity. NaaS contains the optimization of resource
usage and allocations in view of network and computing resources as a cohe-
sive whole.
– HaaS (Healthcare as a service): Infrastructure related to healthcare is delivered
as a service and it can bring significantly cost effective solution for the
healthcare organizations.
– MaaS (Manufacturing as a service): To ensure powerful transformation to auto-
mate various operations as various levels of manufacturing industries, MaaS is
greatly used.
– IoTaaS (IoT as a service): Pioneering companies utilize information potential
available through Internet of Things as a service to sense information and to
create better products for customer satisfaction.
– AaaS (Analytics as a service): Analytics as a Service (AaaS) is a utility based
service that offers a company with the proficiencies of a fully customized analyt-
ics platform for data analysis in cloud [4-9].
72 | Sirisha Potluri, S. Sunaina, P. Neha, Ch. Govind, J. Raghavender, V. MNSSVKR Gupta

Analytics as a
service

Healthcare as a
IoT as a service
service

Cloud computing
technologies

Network as a Healthcare as a
service service

Big Data as a
service

Fig. 4: Cloud computing technologies

7 Types of Cloud Architectures

7.1 Centralized
Clients are tied to the nearest data centres to avoid a high latency in communication
because the cloud resources are partitioned over wide distances geographically.

7.2 Federated
Many small clouds together form a large cloud. It’s very useful when the clients
require a high confidentiality in data when distributing the data geographically.
 A Secure Cloud Infrastructure towards Smart Healthcare: IoT Based Health Monitoring | 73

7.3 P2P cloud

It’s an extension of federated cloud architecture. The only difference is in p2p there
are no centralizing and monitoring components. It consists of independent re-
sources and peers and division of peer’s costs is very low and it leads to minimal
management.
Also in 2010, Koufi et al., have presented a model designed for pervasive access
to cloud Emergency Medical Systems in short form namely EMS [10]. They present
us with a new model of emergency medical systems in a cost effective manner. The
key components of their system is as follows:

– Flexibility to meet ongoing demands 1

– Scalability to adopt dynamic changes
– Interoperability to understand and support system behaviour
– Execution and implementation of platform independent applications and policies: which ensure
easy access and admission to the required data from anywhere in a distributed environment via
any device

The important factors to consider from them are listed above and we want to take all
these into consideration so we propose a model which has a p2p cloud architecture
followed by a combination with BDaas, Saas and Paas. This is complying with all of
the key components in their proposed system. There is scalability since in p2p there
is no centralization and only consists of independent resources and peers. There is
flexibility and interoperability in the services we provide and execution is the most
effective trait of the proposed model. The model emphasizes on the combination of
mobile applications and cloud computing.
Considering all the factors and different approaches provided by various au-
thors and keeping in mind all the challenges are the failures observed in network,
security challenges, and privacy issues of patient information that is being abused
by users, hackers, malware, and etc. Conferring to the analysis and study of the
above articles, many investigators and researchers have focused on reducing pa-
tient response time, increasing safety protection, and speeding up the medical
treatment for patients, we take the help of Paas (Platform as a service), Saas (Soft-
ware as a service) and BDaas (Big Data as a service) alongside P2P and that is our
approach.

8 P2P Architecture
P2P cloud computing combines existing usages and technologies in a more interest-
ing way. There are many security issues regarding user information, credit card
74 | Sirisha Potluri, S. Sunaina, P. Neha, Ch. Govind, J. Raghavender, V. MNSSVKR Gupta

information, etc. Two methods that can help with this are Data security and connec-
tion security. Each storage device in P2P cloud architecture should use data en-
cryption at a higher level.
This technique of storing data can ensure the content level protection such that
stored data cannot be reached by those who have no privilege in accessing it. In a
classic approach, users are directly connected to the servers via a protocol like a
web service.P2P ensures there is a secured connection.
We believe combining P2P architecture and cloud computing is the best solu-
tion for analytics and management of big data. On one hand, P2P networking helps
in decentralization such that peers can control their data and share resources via
clouds. It also provides storage resources, computing and networking required for
big data analytics. Overall architecture of P2P is represented in fig. 5 as shown be-
low.

Cloud user

Node to node
interface

Cloud
interface-
Cloud node
P2PCS
daemon

Fig. 5: Architecture of P2P for cloud platform

 A Secure Cloud Infrastructure towards Smart Healthcare: IoT Based Health Monitoring | 75

9 Security & Privacy

Data privacy makes sure that data is used appropriately and data is available to
those who have authorized access. Privacy can be interpreted as an individual’s
right to determine to what extent information is communicated and how it is com-
municated to others.
The two main classes of data privacy are:
– Data is shared or released with third parties. This includes data collection, pri-
vacy-preserving and data integration data sharing and publishing.
– Privacy-preserving data mining.

A great deal of study has been carried out on the subject of privacy security, and in
recent years there have been many significant review articles reviewing data privacy
taxonomy, privacy protection mechanisms, proposed architectures, and data priva-
cy protection methods; their advantages and disadvantages have been published.
Suggested data privacy security techniques can be categorized into initial ap-
proaches, mathematical analysis, soft computing based on information and ma-
chine learning methods.
Privacy issue is one of the major threats in addition to loss of data, malicious
modification, server crash are some examples of cyber threats such as Yahoo’s three
billion accounts exposure by hackers in 2013.
We proposed CNN based data privacy protection approach to ensure better qual-
ity standards in cloud computing environment. The extraction and inference phases
of the function are divided here and the former takes place locally, while the latter
takes place in the cloud. In this way, during the function extraction stage on the
edge node, confidential information is extracted from the data and data transmis-
sion rates to the cloud are reduced. The privacy-preserving issue in the processing of
data classification is taken into account through the application of deep neural
networks. The CNN (Convolutional Neural Network) algorithm is used here as a deep
neural network. Here, both in the learning and classification processes, a law-
degree non-linear polynomial is replaced to obtain the high precision activation
function. A classical deep neural network (with the ReLU activation function) is
retained to provide data privacy security for the training process, after which the
network changes only when it is used for classification. Combining the above poly-
nomial approximation with batch normalization is the theoretical main innovation
of this work.
In order to effectively solve the data privacy protection problem, encrypted
storage of cloud data is a very outstanding solution. After encryption, the data is
stored in the server provided by the cloud service provider in the form of cipher text,
and in the meantime, the server is also required to return the data to the user when
the user requires it. When the user needs to use the data frequently, it requires a lot
76 | Sirisha Potluri, S. Sunaina, P. Neha, Ch. Govind, J. Raghavender, V. MNSSVKR Gupta

of network bandwidth and user’s time to conduct communication with the server
and realize data encryption and decryption, which will significantly reduce the
usability of cloud computing. In the meantime, after the encrypted data stored in
cloud servers has developed to a certain scale, effective retrieval of encrypted data
has become a new problem that needs to be solved, while the traditional infor-
mation retrieval technology can no long satisfy the requirement of mass data re-
trieval in the cloud storage environment [11-13].

10 Privacy Requirements and Issues in Healthcare

as a Service
The following privacy requirements must be met to guarantee hospital and patient
privacy.

10.1 Privacy of the hospitals’ datasets

Each hospital documents the medical history and diagnoses of patients in its data-
bases, forming a dataset that could be used as a PPRF algorithm training corpus.
However, hospitals usually refrain from sharing them due to the sensitive existence
of such information. Therefore, as we develop our model, their privacy must be
ensured and protected.

10.2 Privacy of the hospitals’ votes

Hospitals may choose to exclude a single or a collection of trees from the ensemble;
however, information from their datasets may be revealed by exposing their votes.
Hence, the votes of hospitals must remain private.

10.3 Privacy of the patients’ diagnosed symptoms

A doctor will use the CDSS to retrieve potential diagnosis based on the patient’s
symptoms in order to diagnose a patient. If no assurances of privacy were provided,
however, no patient would allow his or her symptoms to be fed into the system. The
symptoms of patients must, therefore, remain private.
 A Secure Cloud Infrastructure towards Smart Healthcare: IoT Based Health Monitoring | 77

11 Cloud Services to Preserve Data Security in

Healthcare
Preserving Data security is needed for the hour especially when cloud storage is
used by various organizations of different fields ranging from educational institu-
tions and huge MNCs relying on services like google drive to individual users like
Dropbox, Box, Amazon Drive, Microsoft OneDrive etc. Such users trust these ser-
vices and want to maintain their data’s privacy by not letting it get corrupted or not
losing it by any means. And if this trust continues millions of other users might want
to store data on the cloud. Stored cloud data needs to be decrypted before a 3rd
party could breach information since it is stored in an encrypted form. Adding to
this. These are various methods by which users can inculcate their own security
measures in addition to what is provided by the system. All things considered, there
are still holes among practices and proposed arrangements, irreconcilable circum-
stances, and difference on prerequisites and ideas.
Comparison of cloud security and privacy preserving algorithms which are used
in healthcare is represented in table 2 [14-21].

Tab. 2: Comparison of cloud security and privacy preserving algorithms which are used in healthcare

S.No Model/Algorithm Parameters

1 Bilinear pairing cryptography Cost, time

2 Identity Based Encryption Membership, identity
3 Homomorphic Encryption Patient’s data like clinical records, scanning
and x-ray images
4 Attribute Based Encryption (ABE) Security parameter, user data
5 Multi-Objective Privacy-Aware workflow Storage cost, time, data,
scheduling algorithm(MOPA)

12 Real Time Healthcare Dashboard- A Case Study

Based on Blood Pressure Readings
Real time healthcare dashboard for tracking a patient health condition such as
blood pressure readings, heat beat monitoring, blood sugar observing, cancer cell
growth perceiving etc. is essential. Based on which we can perform time series anal-
ysis and go for a better prediction. The following code shows real time health care
dashboard to monitor a patient.
78 | Sirisha Potluri, S. Sunaina, P. Neha, Ch. Govind, J. Raghavender, V. MNSSVKR Gupta

Listing 1: Real time healthcare dashboard for tracking a patient health condition- blood pressure
readings

# Real time healthcare dashboard for tracking a patient health condition

based on blood pressure readings
def _health_monitoring():
arr_results = []
client_record = WithingsApi(creds)
health_readings = {
# analyse readings based on past history
'past_reading': {
'x_p': '',
'diastolic_p': '',
'systolic_p': '',
'pulse_p': '',
's_m_average_p' : {
'diastolic_p': '',
'pulse_p': '',
'systolic_p': '',
}
}
}
# Get patient readings
p_measures = client_record.get_measures()
# Get patient last reading date
l_r_date = p_measures[-1].date
p_counter = 1
# Get patient raw readings from the record
raw_readings_p = {
'systolic_p': [],
'diastolic_p': [],
'pulse_p': [],
}
# Extract patient measure
for p_measure in p_measures:
if p_measure.systolic_p_blood_pressure\
and p_measure.diastolic_p_blood_pressure:
next_pdate = l_r_date + timedelta(days=counter)
# sort the values of date times
health_readings['past_reading']['x'] += '"' +
p_measure.date.strftime('%Y/%m/%d %H.%M.%S') + '",'
 A Secure Cloud Infrastructure towards Smart Healthcare: IoT Based Health Monitoring | 79

health_readings['future_reading']['x'] += '"' +
next_pdate.strftime('%Y/%m/%d %H.%M.%S') + '",'
health_readings['past_reading']['systolic_p'] +=
str(p_measure.systolic_p_blood_pressure) + ','
health_readings['past_reading']['diastolic_p'] +=
str(p_measure.diastolic_p_blood_pressure) + ','

raw_readings_p['systolic_p'].append(p_measure.systolic_p_blood_pressure)

raw_readings_p['diastolic_p'].append(p_measure.diastolic_p_blood_pressur
e)
if p_measure.heart_pulse and p_measure.heart_pulse>33:
raw_readings_p['pulse_p'].append(p_measure.heart_pulse)
health_readings['past_reading']['pulse_p'] +=
str(p_measure.heart_pulse) + ','
p_counter += 1
return health_readings

13 Conclusion
Based on the investigations observed in the literature, the performance and enact-
ment of telemedicine service through hosting and presenting options and choices of
Amazon EC2. The efficiency and proficiency is examined and studied by answering
the number of requests per seconds. According to the description provided by them,
dynamic resource provisioning on web tier with medium type instances is much
better than static allocation with large and extra-large instances. A feasible solution
is obtained but it’s not implemented in real time.
Cloud-based homecare cloud model is best suited for patients with respiratory
diseases. They have a conscious platform that can dispense and store a patient’s
respiratory data alongside caregivers, families, and infirmaries quickly. By utilizing
mobile technologies and models, the saturation and capacity of oxygen can be mon-
itored and examined and the information can be distributed to patients and medical
specialists.
In P2P there are no centralizing and monitoring components. It consists of in-
dependent resources and peers and division of peer’s costs is very low and it leads
to minimal management. The PaaS provider will fund much of the infrastructure
and other IT related services, which users can access anywhere in a distributed
environment via a web browser it is affordable as well PaaS offerings are usually
used for mobile apps , cross platform apps and devops tools. SaaS and BDaas will
help in the all-round development of cloud computing in healthcare.
80 | Sirisha Potluri, S. Sunaina, P. Neha, Ch. Govind, J. Raghavender, V. MNSSVKR Gupta

14 References
[1] Rajkumar B, Chee Y, Srikumar V, James B, Ivona B, Cloud computing and emerging IT plat-
forms: vision, hype, and reality for delivering computing as the 5th utility, Future Generation
Computer Systems, 2009, Volume 25, Issue 6, Pages 599-616, ISSN 0167-739X,
https://doi.org/10.1016/j.future.2008.12.001
[2] Peter M, Timothy G, The NIST definition of cloud computing, National Institute of Standards
and Technology: Special Publication (NIST SP), 800-145, 2011
[3] Al-Issa Y, Ashraf O, Tamrawi A, eHealth cloud security challenges: a survey, Journal of
Healthcare Engineering, 2019, vol. 2019, Article ID 7516035, 15 pages,
https://doi.org/10.1155/2019/7516035
[4] Harika K, Manisha G, Potluri S, An IoT based solution for health monitoring using a body-worn
sensor enabled device, Journal of Advanced Research in Dynamical and Control Systems
(JARDCS), 2018, Volume 10, Issue 9, PP: 646-651, ISSN 1943-023X
[5] Potluri S, Avinash M, Health record data analysis using wireless wearable technology device,
Journal of Advanced Research in Dynamical and Control Systems (JARDCS), 2018, Volume 10,
Issue 9, PP: 696-701, ISSN 1943-023X
[6] Avinash M, Potluri S, A study on technologies in cloud-based design and manufacturing, Inter-
national Journal of Mechanical and Production Engineering Research and Development
(IJMPERD), 2018, Volume 8, Issue 6, PP: 187-192, ISSN(P): 2249-6890, ISSN(E): 2249-8001
[7] Potluri S, Achyuth S, Elham T, Mohanty S N, IOT enabled cloud based healthcare system using
fog computing: a case study, Journal of Critical Reviews, 2020, ISSN- 2394-5125, Vol 7, Issue 6,
PP: 1068-1072, doi: 10.31838/jcr.07.06.186
[8] Goyal S, Public vs private vs hybrid vs community - cloud computing: a critical re-
view, International Journal of Computer Network and Information Security, 2014, 6, 20-29
[9] Jansen W, Grance T, Guidelines on security and privacy in public cloud computing, NIST Special
Publication, 2011, 800-144, pp. 5
[10] Koufi V, Malamateniou F, Vassilacopoulos G, Ubiquitous access to cloud emergency medical
services, Proceedings of the 10th IEEE International Conference on Information Technology and
Applications in Biomedicine, Corfu, Greece, 2010, pp. 1-4, doi: 10.1109/ITAB.2010.5687702
[11] Haider W, Iqbal W, Bokhari S, Bukhari F, On providing response time guarantees to a cloud-
hosted telemedicine web service, Zhang Y, Peng L, Youn CH, Cloud Computing, CloudComp
2015, Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommu-
nications Engineering, 2015, vol 167, Springer, Cham, https://doi.org/10.1007/978-3-319-
38904-2_24
[12] Risso A, Neyem A, Benedetto J, Carrillo M, Farías A, Gajardo M, Loyola O, A cloud-based mobile
system to improve respiratory therapy services at home, J Biomed Inform, 2016, 63, 45-53, doi:
10.1016/j.jbi.2016.07.006
[13] Navimipour J, Navin H, Rahmani M, Hosseinzadeh M, Behavioral modeling and automated
verification of a cloud-based framework to share the knowledge and skills of human resources,
Computers in Industry, 2015, 68, 65–77
[14] Neghabi A, Navimipour J, Hosseinzadeh M, Rezaee A, Load balancing mechanisms in the soft-
ware defined networks: a systematic and comprehensive review of the literature, IEEE Access,
2018, 6, 14159–14178
[15] Akinsanya, Opeoluwa O, Papadaki M, Sun L, Current cybersecurity maturity models: how effec-
tive in healthcare cloud?, CERC (2019)
 A Secure Cloud Infrastructure towards Smart Healthcare: IoT Based Health Monitoring | 81

[16] Duncan B, Pym D, Whittington M, Developing a conceptual framework for cloud security assur-
ance, IEEE 5th International Conference on Cloud Computing Technology and Science, Bristol,
UK, 2013, pp. 120-125, doi: 10.1109/CloudCom.2013.144
[17] Kumari S, Kamal A R, Optimal integrity policy for encrypted data in secure storage using cloud
computing, Indian J Sci Technol, 2016, vol.9, no.8, 2016, pp.1–10, 2016
[18] Monikandan S, Arockiam L, A security service algorithm to ensure the confidentiality of data in
cloud storage, Int J Eng Res Technol, 2014, 3(12), 1053–1058
[19] Shabbir A, Shabbir M, Rizwan M, Ahmad F, Ensuring the confidentiality of nuclear information
at cloud using modular encryption standard, Security and Communication Networks, 2019, vol.
2019, Article ID 2509898, 16 pages, https://doi.org/10.1155/2019/2509898
[20] Singh N, Singh A, Data privacy protection mechanisms in cloud, Data Sci. Eng, 2018, 3, 24–39,
https://doi.org/10.1007/s41019-017-0046-0
[21] Delettre C, Boudaoud K, Riveill M, Cloud computing, security and data concealment, Proceed-
ings, 16th IEEE Symposium on Computers and Communications, Kerkyra, Greece, 2011, 424
431, 10.1109/ISCC.2011.5983874 2-s2.0-80052768216
I. Indrani, SVB Revanth, S. Akhil Durga, Sirisha Potluri, Sachi
Nandan Mohanty
Internet of Cloud: Secure and Privacy Preserving
Cloud Model with IoT Enabled Service
Abstract: Security and safety are among the critical difficulties of Internet of Things
(IoT). Updating inappropriate gadgets, the lack of productive and powerful security
meetings, customer ignorance, and viewing a well-known powerful gadget are some of
the problems IoT faces. In this work, we explore the basis of IoT frameworks and security
efforts, and we see other safety and security matters, methods used to verify IoT-based
components and structures, security arrangements in place, and the best protection
models are important and suitable for the various layers of IoT-operated applications. In
this work, we have proposed another embossed IoT model: standard and expanded with
security and safety components and layers of physical evidence. The proposed
cloud/edge IoT framework is developed and tested. The background below spoke to IoT
centres created via Amazon Web Service (AWS) such as Virtual Machines. The centre
layer (edge) has been developed as a resource unit for Raspberry Pi 4 with the help of
Greengrass Edge Environment in AWS. We used cloud-enabled IoT weather in AWS to
create the top layer (cloud). Security meetings and basic management meetings were
among all these areas to ensure the security of customer data. We have verified security
announcements to allow data to flow between layers of the proposed cloud-enabled
model. Not only does the proposed framework model eliminate the perceived security
vulnerabilities, but it can also be used alongside the best security measures to combat
network security risks facing all the latter layers; cloud, edge, and IoT

Keywords: Internet of Things, Cloud Computing, Cloud Privacy, Cloud Security,

Edge Computing

||
Indrani Inapakolla, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana
501203, India, [email protected]
SVB Revanth, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI Foundation
for Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana 501203, India,
[email protected]
Siliveru Akhil Durga, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana
501203, India, [email protected]
Sirisha Potluri, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana
501203, India, [email protected]
Sachi Nandan Mohanty, Department of Computer Engineering, College of Engineering Pune,
Wellesley Rd, Shivajinagar, Pune, Maharashtra 411005, India, [email protected]

https://doi.org/10.1515/9783110732573-005
84 | I. Indrani, SVB Revanth, S. Akhil Durga, Sirisha Potluri, Sachi Nandan Mohanty

1 Introduction
Internet of Things (IoT) points to the concept of articles related to gadgets of many
kinds on the Internet or wireless Internet. The popularity of IoT or Internet of Things
has grown rapidly, as these new devices are used for a variety of purposes, includ-
ing literature, travel, education and business improvement. IoT introduced the con-
cept of connectivity, which means that organizations and individuals can communi-
cate from far and wide easily. Kevin Ashton coined the term ‘IoT’ in 1999 to promote
the concept of Radio Frequency Identification (RFID), which combines sensors and
actuators. Nevertheless, the first idea was introduced in the 1960’s. At the time, the
idea was inevitable to register or install the Internet. Ashton presented the IoT’s idea
of promoting good exercise with chains. Be that as it may, the versatility of IoT has
helped to boost strong availability during 2010. The Chinese government gave
greater demand to the IoT by introducing a five-year plan. Approximately 26.66
billion IoT device resides in the rotating earth. The big bang started in 2011 with the
introduction of home appliances, wearable gadgets, and sharp energy meters. The
rapid explosion of IoT has helped organizations and in various ways improved sta-
tistical and business research methods. In addition, IoT has improved the way peo-
ple live by introducing computer management. However, uncontrolled explosions
have exacerbated security and safety challenges [1].
The unrecognized use, not the emergence of passwords, and the absence of up-
dates to gadgets have increased the risks of online security and the acceptance of
malicious applications on sensitive IoT frameworks. Such improper training in-
creases the likelihood of information leaks and accidents. The vast majority of secu-
rity experts view the IoT as a vulnerable area for digital attacks due to weak security
meetings and arrangements. Unless a few security items are designed to protect IoT
gadgets from digital attacks, security rules are not properly documented. In line
with these lines, end clients have not been able to use defensive tactics to destroy
information attacks. The developers of the program have created a variety of mal-
ware to contaminate IoT gadgets from the night before 2008. They are organizing
various forms of crime to steal sensitive information in order to arouse public repre-
sentatives to share sensitive information. Similarly, workplaces and certain compa-
nies are constantly facing security breaches due to prominent attacks. If gadget
manufacturers and security experts assess digital risks accurately, they can create a
protective component to prevent or kill digital accidents [2].
IoT enabled gadgets used for automated applications and for various business
purposes. Applications help these organizations achieve their rivals. However, due
to the unnecessary selection of various savvy gadgets that share information and
connectivity, access to security and information becomes a major concern for many
organizations, as it interferes with workflow, exercise, and organizational manage-
ment. It is important for professionals to overcome these dangerous concerns and
 Internet of Cloud: Secure and Privacy Preserving Cloud Model with IoT Enabled Service | 85

make comprehensive safety efforts and strategies to protect the resources of their
business and ensure the continued integrity and integrity of management. For ex-
ample, powerful IoT kitchen appliances associated with a local organization can be
a source of entry for planners to gain business acceptance and in addition to rele-
vant information or to control and disrupt the business process.
New developments are emerging, or changes are being made in existing ones.
Consider the latest developments in the 5G organization, for example. 5G is required
to take a basic role in IoT frameworks and applications. It is upright enough to be
recognized and is interested in the safety and security risks you can imagine, with
its high frequency and data transfer. However, the shorter frequency necessitates
adjustment at the base, as a result of which the need for additional channels to cov-
er the same region is covered by new different remote techniques [3].
In this work, we plan to provide a framework for IoT applications, benefits, and
expected risks. In addition, building a structure that needs to be considered and further
enhances the best practice by doing or researching existing programs or developing
new ones. Based on the findings, we offer suggestions for avoiding such dangers and
treating potential safety concerns. This work will direct management organizations to
continue to implement strategies, instruct storage clients and objects, and IoT-related
partners to create and implement appropriate security and protection measures.
We developed our model using Amazon Web Service (AWS) as a confirmation of
ideas, which later referred to real sensory frameworks that mimic the overall IoT
structure. By creating a draft, we can post and consider different security measures
by creating real locations and benchmarks.
We have embraced account research philosophy to explore a set of experiences
and the basis of IoT frameworks, their security and security issues, and comparative
measures. We have expressed our opinion about the inclusive and expanded IoT
model and its protection and security. We designed and tested the cloud/elevated
IoT model with the optical device (sensors), the hub edge (Raspberry Pi), and cloud
management (AWS). This provision was intended to test the model we have pro-
posed in the sections accompanying this paper. Our work does not give cunning to
different IoT applications (well-being, intelligent urban communities, good chain,
transport, etc.); their highlights, their preferences, and their challenges, or security
risks that can be considered within these applications. The text is rich in such a
thing. In this work, we wanted to have a complete review of the evidence and to lay
the groundwork for further testing and evaluation.
The rest of the paper is organized as follows: the next section introduces
bookkeeping followed by IoT security and protection challenges. In Section 4, we
examine the end of the Internet of Things at the end. Section 5 introduces the pro-
posed cloud-based/edge IoT models: standard and extended with protective and
security components and layers of visual evidence. This section similarly illustrates
the use of the proposed AWS cloud model and the conditions on the edges and
package of the Raspberry Pi 4. Section 6 closes this function.
86 | I. Indrani, SVB Revanth, S. Akhil Durga, Sirisha Potluri, Sachi Nandan Mohanty

2 Review of Literature
The creators have pointed out that there are various problems, for example, sticking
to and emphasizing attacks and other unauthorized access, which has reduced re-
spect for customer information. There are potential programs that can assist a per-
son in making various security measures that can help ensure their IoT gadgets. As
mentioned, various security risks have arisen at the moment, and they can access
IoT Technologies with their integrated organization. Organizations should provide
testing and filtering tools to all IoT gadgets that can detect any type of risk identified
safely and try to eliminate the risk of intrusion. Trainers and traffic analysts help
identify and evaluate various digital risks [4].
Many administrators have introduced part of the difficulty or loading of widgets
to various IoT gadgets and its porters. The different reproductive tools, models, and
accessibility of the various components that can ensure this security meeting can
also help to create a well-defined IoT security meeting for the novel. Most would
agree with the rapid development of the IoT security tests and the various simula-
tion tools as long as the actors supported this experiment. If IoT gadgets are likely to
attack, then the problems will get worse.
The creators acknowledge that, while there are great benefits to clients for the
Internet of Things, there are joining forces that should be taken seriously. Online
security and security opportunities are the most important issues discussed. These
two represent the tremendous difficulties in some business organizations as an open
organization. The usual attacks of network protection have shown the weakness of
new IoT devices. This is simply because the interaction of organizations on the In-
ternet of Things brings openness to the anonymous and unreliable Internet, which
requires security arrangements for novels. Also, it is important to emphasize the
important norms and standards of the IoT Cyber Security framework regarding the
creation of an IoT security framework. As noted, one of the key steps to consider is
the conclusion of an agreement with a variety of gadgets with different book confer-
ences. Differences in meetings prevent separate service contracts from operating
and are essential elements to be found in the security structure of each Internet of
Things network. He pointed out that to ensure the reliability of the IoT system in the
online security sector, little progress should be made to help reduce the security
complexity of the IoT network. Likewise, the founders have shown that doing more
things is an important part of the success of the Internet of Things network security
program. The researchers said the IoT climate should work with a variety of factors
to address the billions of Internet-related challenges and network security. Accord-
ingly, the journal has suggested that the weather protection of the IoT network
should also keep testing, for example, installation testing, partial testing, frame-
work testing, and consistency testing, adequately minimizing complications and
 Internet of Cloud: Secure and Privacy Preserving Cloud Model with IoT Enabled Service | 87

risks. In a similar vein, the creators have introduced part of the current IoT network
security systems [5].
Some important safety efforts have been made by the provider, and have shown
that it does not prepare the provider to make better arrangements. Due to the online
protection of the Internet of Things, organizations will probably not make the nec-
essary arrangements. Also, the creators are introducing real-time flexible and digital
frames as a whole, from mechanical control frameworks, current vehicles to the
basic framework. Recent features and functions, for example, Industry 4.0 and In-
ternet of Things (IoT), ensure the innovation and innovation plans of new customers
through a strong network and the effective use of the new age of installed gadgets.
These frameworks produce, rotate and trade a wide range of functional infor-
mation. Security and the secretive beliefs that make digital attacks an attractive goal
of the Internet of Things cause real corruption and disrupt people’s lives. Network
security and security are important on the grounds that they may pose a risk. The
multiplicity of these frameworks and the potential consequence of digital attacks
present new dangers in today’s related IoT frameworks. Potential answers to the
security and safety challenges are the standard security frameworks of modern IoT
frameworks. The current IoT frameworks have not been adequately developed to
ensure appropriate capabilities.
As such, there has been a remarkable increase in the investigation and evalua-
tion of various security issues on the IoT. One of the main objectives of IoT security
is to protect, maintain, and ensure that each client can develop the assurance,
foundations, and assurance of access to the different management provided by the
IoT biological system. Similarly, alternative IoT security testing enhances basic
capabilities with the help of various recreational tools such as multi-component
computers [6].

3 IoT Security and Privacy Issues

IoT has brought clients great benefits; however, a few difficulties joined in. Network
security and privacy prospects are important concerns for analysts and security
professionals in question. These two represent the great difficulties of certain busi-
ness organizations such as open organizations. Outstanding network attacks have
highlighted the weaknesses of IoT development. This weakness lies in the fact that
organization’s connections to the Internet of Things bring access to an anonymous
and unreliable Internet that requires security arrangements for novels.
In the vast majority of known problems, none of them affect IoT variations, for
example, security and protection. However, it is unfortunate that clients do not have
the necessary assurance of security impacts until after the break has occurred, caus-
ing serious damage, for example, loss of important information. With the advent of
88 | I. Indrani, SVB Revanth, S. Akhil Durga, Sirisha Potluri, Sachi Nandan Mohanty

advanced security that has undermined customer protection, consumer’s craving

for helpless security is now diminishing. In an ongoing study aimed at security and
safety, the Internet of Things customer service did not go well. There has been a ton
of weakness in the current auto industry [7-8].

3.1 Safety
IoT is integrated from standard PCs and gadget registrations, making it incapable of
dealing with security challenges of various types:
– Many Internet of Things gadgets are designed for large size layouts.
– Typically, IoT configurations contain a number of similar or almost indistin-
guishable resources with comparable features. This similarity enhances the
magnitude of any security vulnerabilities that could completely affect a large
number of them.
– This advancement means that the potential number of connections between IoT
gadgets is staggering. It is even more certain that most of these gadgets can set
up organizations and talk about different gadgets as a result in an unexpected
way. This needs to be considered with the available devices, methods, and
techniques available for IoT security.

Indeed, while the issue of data security and innovation is not a new phenomenon,
the use of IoT has brought about different measures to lean towards. Consumers
need to rely on Internet technology and management is safe from mistakes, espe-
cially as these innovations continue to gain more time and are incorporated into our
normal days of life. With partially enhanced IoT devices and management, this is
one of the major routes used for digital attacks such as the importation of customer
information by leaving information streams unsatisfactory.
The concept of IoT gadget connectivity means that if the gadget is poorly veri-
fied and connected to it has the potential to influence global security and flexibility
around the world. This behaviour is achieved by testing a large IoT gadget function.
Aside from the ability of certain gadgets to have the option to interact directly with
different gadgets, it means that clients and developers of IoT are all committed to
ensuring that they do not expose separate clients such as the Internet itself from
potential damage. The standard approach required to build a robust and efficient
response to a crisis is as seen now in IoT [9-10].
In terms of validation, for example, IoT faces a different weakness, which re-
mains a major problem in security systems in many systems. The authentication
used is limited by the way it protects against just one risk, for example, Denial of
Service (DoS) or retaliation. Data security is one of the most critical regions in IoT
authentication due to the increase in unsafe use due to their diverse IoT climate
information. If possible, for example, we have taken a picture of Visa contacts.
 Internet of Cloud: Secure and Privacy Preserving Cloud Model with IoT Enabled Service | 89

These cards are equipped to allow card numbers and names to be used without IoT
verification; this makes it possible for program planners to have the opportunity to
purchase merchandise using the cardholder’s financial balance number and their
personality.
Another notable attack on IoT is the man in the middle, where an outdoor pho-
tography channel is shown to photographing characters of large areas involving
network trading. The man, at the centre of the attack causes the bank employee to
see the exchange as a legal term because the enemy does not need to know the
character of the suspect.

3.2 Secret
The concept of easy-to-use IoT depends on how well it looks at people’s safety pref-
erences. Concerns about the safety and potential harm of joining an IoT may be
crucial in maintaining full IoT acceptance. It is fundamental to see that customer
protection and management rights are critical to ensuring customer confidence and
confidence in the Internet of Things, compatible gadget, and related advertising
management. A lot of work is trying to get the assurance that IoT is reconsidering
security issues such as increased monitoring and tracking. The explanation behind
the concern for security is the direct result of the inevitable understanding of the
complexity of the past where the cycle of testing and dissemination of data on IoT
can be done almost anywhere. Universal access to the Internet is equally a funda-
mental figure that helps to understand this problem on the grounds that unless
there is an exciting system, then it will be explicitly acceptable to access individual
data from any aspect of the earth [11].

3.3 Cooperation
A separate climate for IoT specialization is known to suppress customer motivation.
Apart from the fact that full integration is rarely practical in real estate and man-
agement, clients are reluctant to purchase and manage items when there is no flexi-
bility and concern about getting inside the vendor. Inadequate IoT naming can
mean that there will be a negative reversal of the assets of the management systems
we work with.
Cryptography is one of the most widely used cornerstones to provide security in
many ways. A powerful system of self-defence against potential attacks is not con-
sidered using a single security system. This, in turn, requires different levels of pro-
tection from the risks of IoT authentication.
By improving the improved safety points and adding these excellent features,
hacks can be prevented. This avoidance is due to the fact that clients will purchase
90 | I. Indrani, SVB Revanth, S. Akhil Durga, Sirisha Potluri, Sachi Nandan Mohanty

items that currently have legal security that exposes the preventing vulnerabilities.
Online security programs are part of the measures in place to ensure IoT is secure.
Alternatively, features and minor concerns may affect the negotiation efforts of
the Internet of Things gadgets; this includes:
– Periodic updates: for the most part, IoT manufacturers are updating security
updates on a quarterly basis. OS forms and security patches are also redesigned.
In this way, program planners get ample opportunity to break the security con-
ventions and take the details involved.
– Embedded passwords: IoT gadgets store embedded passwords, enabling help
professionals to investigate OS issues or deliver important updates remotely. Al-
ternatively, programmers can use the feature to securely access gadgets.
– Automation: regular, performance and end customers using computer equip-
ment for IoT social media frameworks or to improve business practices. Alterna-
tively, if retaliatory areas are not displayed, the installed AI can access those re-
sources, which will allow risks to enter the framework.
– Remote access: IoT gadgets use various organizational conferences for remote
access such as Wi-Fi, ZigBee, and Z-Wave. Often, clear limitations are not ad-
dressed, which can be used to deter cybercriminals. Therefore, program plan-
ners can quickly establish a revenge organization through these remote access
conferences.
– Several integration of external applications: a number of application programs
are available online, which can be used by organizations to perform specific
tasks. In any case, the functionality of these applications was not detected
without a problem. In the event that final clients and representatives present or
access such programs, risk specialists will enter the framework and damage the
database.
– Incorrect gadget authentication: most IoT applications do not use authentica-
tion management to block or limit network risks. In this way, the attackers enter
the entrance and undermine security.
– Weak Device Recognition: In general, all IoT manufacturers set up identifiable
gadget identifiers to monitor and track gadgets. In any case, few manufacturers
do not keep their safety strategies. After all, following questionable internet
tests becomes very dangerous [12-16].

4 The Future of IoT

At the moment, objects and structures are enabled for network access and have the
ability to register to talk about compatible gadgets and devices. Increasing the ca-
pacity of the organization in all areas of thought will make our lives more successful
and help us set aside time and money. Besides, connecting to the Internet more and
 Internet of Cloud: Secure and Privacy Preserving Cloud Model with IoT Enabled Service | 91

more aims to talk about the dangers of digital. Web-authorized content becomes the
responsibility of cyber criminals. The expansion of the IoT market increases the
magnitude of the potential risks, which could affect the profitability and well-being
of gadgets and therefore our security. Reports include the frequency of information
breaks that are definitely extended since 2015; 60% in the USA only. A review of all
that has led to Japan, Canada, the UK, Australia, the USA and France found that
63% of IoT consumers thought these gadgets were a threat due to poorly regulated
security. The test findings also included that 90% of customers are unsure about IoT
network security.
Momentum research has investigated various creative processes to measure dig-
ital attacks and enhanced security arrangements. Part of the experimental detected
arrangements are recorded below; Transfer encryption processes: authorizing ro-
bust and robust encryption methods can extend online protection. Encryption as-
sembly performed on both cloud and gadget environments. As a result, the editors
of the program were unable to comprehend the intangible and abusive data struc-
ture. Consistent assessment of emerging risks: safety risks are regularly assessed.
Such groups analyse the impact of IoT threats and develop effective control
measures through continuous testing and evaluation.
Adding duplication of updates: gadget manufacturers should grow fewer fixes
than counter important updates. Such a procedure can reduce the uncertainty of
starting the adjustment. Also, continuous updates will help clients by misleading
dangerous digital assets from various sources.
Transfer solid gadget test devices: a large part of the new test is designed to
make heart-focused gadget viewing strategies so that those questionable tests can
be tracked and controlled without any problem. Many IT organizations are familiar
with a gadget capable of viewing hazardous materials. Such devices are very im-
portant in risk assessment, which helps organizations to build modern control tools.
Create archived customer rules to create security assumptions: a large part of
the break breaks and IoT attacks occur due to a lack of customer thinking. For the
most part, IoT security efforts and regulations are not identified when clients pur-
chase these gadgets. In the event that gadget makers explicitly determine potential
IoT risks, clients may end up with these issues. Organizations can similarly organize
preparatory projects to improve security awareness. Such projects guide clients to
build strong passwords to update them consistently. Also, clients are told to restart
security adjustments consistently. Clients add more and are said to maintain a range
of strategies from spam messages, external applications, or sources that can sell IoT
security.
There will be more than 30 billion IoT gadgets by 2025. Before that, people knew
about IoT work, yet they discarded the idea by looking at how confusing the idea
looks and how tempting it is to do it. However, with the development of new inven-
tions, it is now possible for people to believe that this is only possible as the level of
development of IoT increases the new steps step by step. In 2020 and earlier, for
92 | I. Indrani, SVB Revanth, S. Akhil Durga, Sirisha Potluri, Sachi Nandan Mohanty

example, intelligent home controls and bright lighting are just a few examples of
how IoT is used to protect energy and reduce tariffs and this can add to the amazing
motivation of why so many people choose IoT gadgets.
Many urban communities will shine. There will be better traffic management;
roads will be freed from closures, urban areas will benefit from reduced pollution,
safety will be the high expectations for this with a massive IoT massacre.
Management of medical care is very expensive, with the majority of diseases
persisting in ascendancy. We are heading for a time when essential medical services
can be integrated to find other people, especially as people are more prone to infec-
tion. Other than that, apart from the fact that the new product is not designed to
prevent people from growing, it could help make medical services easier on the
pocket to the point of availability. For example, by delivering regular clinical checks
from an emergency clinic to a patient’s home, this will be of great help to patients.
Ongoing testing using Internet-related gadgets is one of the ways in which it will
help save the lives of many patients. Timely alarms are critical in case of emergency
situations, the same number of IoT gadgets in clinics will continue to be associated
with the most important details for continuous follow-up. Patient satisfaction will
be fundamentally improved.
In this work, we propose another perspective on IoT models: inclusive and ex-
tended with security and security components and ID and team layers. We have
designed a cloud/raised edge of the IoT framework to make the proposed IoT mod-
els. Therefore, in this work we begin by introducing inclusive and expanded models,
at which point we demonstrate our experimental design and climate (use of a con-
centrated model), and then present and discuss the results and findings [17].

5 Generic IoT Layers and Data Integration Model

The standard design of the IoT model, from the creator’s point of view, is uncertain
as to whether there are any comparable observations in writing, as shown in fig. 1,
containing gadget, cloud and end customer layers. The gadget layer consists of a
pool of Internet-enabled sensors, data acquisition hardware, and book conferencing
to send information to the nearest or remote capacity for additional preparation.
These gadgets allow the client to collect data on an ongoing basis through various
purchase waves. Cloud layer contains data collected from additional management
sensors, audio output, including output, and data scripting. This information is
later regarded as a network of emotional selections that conducts the investigation
of complex data and man-made knowledge to provide an option in relation to hu-
man well-being. The final client layer, which consists of the receiving client, can be
in various frameworks. Of concern are smart gadgets, where safety and security
challenges exist. Within the boundaries of these three layers, a rundown of sublay-
 Internet of Cloud: Secure and Privacy Preserving Cloud Model with IoT Enabled Service | 93

ers or modules is installed to ensure the strength of the air support network. To
ensure that data is sent and managed quickly to provide basic options that will not
stand until the data is sent to the cloud, we present the edge-finding capabilities
that can stay in such a great option, and simultaneously save duplicate information
and send it to cloud layer to prepare and store long-term retrieval. In some cases, we
need to send orders or references to certain wearable gadgets to update their pur-
chase or usability ratio, and this will require alternative assembly and security
measures.
Fig. 2, shows the extended extension of the excluded model. We are seeing an ex-
pansion of new layers; on the brink of misery. These two layers can overcome the
issues of inactivity from the dependence of the benefits of the cloud layer and can be
resolved in a quick selection. Edge registration occurs on gadgets where the sensors
are plugged into or actually turned off. They provide continuous selection and control
over the sources of information, and at the same time, discuss the various layers to
convey information by integration, remote, and investigative. The haze enrolment
layer motivates edge-testing experiments on all the most notable intellectual property
associated with neighbours and is actually removed from the nerves and sources.
These additional benefits create significant safety and security challenges [18].

6 Security and Privacy Policies

Cloud-based management is often regarded as the basic IoT framework that pro-
vides information storage, data preparation and information sharing. Program
planners and abusers focus on IoT processing gadgets and on hubs that store or
transmit sensitive information. For example, tolerable data and electronic clinical
records make the medical services framework an important part of the program.
Each layer of the IoT model presents security challenges and, at the same time, the
opportunity to maintain safety and security practices and meetings. For example, in
the gadget layer, sensor information is sent to the edge, in the haze, and then to the
cloud, the requirement of approval and announcements that trust explicit workers
to mitigate these attacks. Firmware and machine address security ensures and that
is the only snow, however, this comes as a result of power consumption, as part of
remote-enabled gadgets, for example, wearable battery-powered. Those safety ef-
forts must be revived to achieve both safety, and power [19].
94 | I. Indrani, SVB Revanth, S. Akhil Durga, Sirisha Potluri, Sachi Nandan Mohanty

End User

Device/ Privacy and Security policies

Sensors

Cloud

Fig. 1: Internet of Things (IoT) generic model with privacy and security policies

On the cloud layer, security efforts need to ensure organizational meeting between
edge areas and mass areas and periodically from sensors. Converting meeting mes-
sages, highlighting encryption points, and enabling all provides minimal infor-
mation for check-in and login. At the data management and final client level, we
need to ensure that data retention and retrieval of fixed data is protected from SQL
attacks, smells, and attacks of sensitive identity theft documents, providing renew-
able support and following HIPPA procedures (in social frameworks) [30) ]. The
integration of the information can familiarize the department with the programmers
to isolate the client, as a result of which security comes in. As IoT gadgets can join
and leave the organization of sensors and sources, this can add a lot of confusion to
the general security effort strategies, hence the need for new smart and diverse se-
curity efforts [20].
 Internet of Cloud: Secure and Privacy Preserving Cloud Model with IoT Enabled Service | 95

End user layer

Smart devices, stations ...

Cloud compuitng
layer
Edge computing
Fog compuitng Data storage,
Device Layer layer
layer feature
Multiple sensors Local data and extraction,
Multi- Sensor
and data sources real time anamoly
data fusion
analysis detection,
decision system
and control

Fig. 2: Cloud Edge IOT model

7 Implementation of the Recommended Cloud-Edge-

IoT Model
Our approach to ensuring security measures put in place before sending enabled
gadgets to IoT has been proven in terms of planning and ensuring that they can
safely transfer and share information, protecting data encryption. Fig. 3 below
shows the emergence of machinery, system, and book model. The model contains
AWS cloud as expert cloud, Raspberry Pi 4 as Edge Node, and Virtual Machines as
IoT gadgets. The framework we have created is a paid AWS record for full ac-
ceptance of assets provided by AWS, including testament keys and encryption, ap-
proval and validation.
From the assets acquired by AWS, we have implemented AWS Identity and Ac-
cess Management (IAM) web administration. It allows us to control client login by
setting up an IAM (customer representation) account for all clients. For security
reasons, we did not use the AWS Root account but started the IAM client with au-
thorized authorization. Designing the Raspberry Pi as an Edge hub (AWS Green
grass Core), AWS Green grass Core works directly with the cloud and works locally.
96 | I. Indrani, SVB Revanth, S. Akhil Durga, Sirisha Potluri, Sachi Nandan Mohanty

The Raspberry Pi was designed with an outstanding addition to the Linux durabil-
ity. Creating a partnership between AWS and Raspberry. We used the AWS Green
grass Core to create a circle with the centre gadget, as well as any remaining IoT
gadgets to allow them to speak on the edge.
We need announcements to verify all gadgets with AWS. We have issued rec-
ommendations, private and public keys for safe and secure relations with AWS.
Institutional certification is generated by AWS once we have done the Greengrass
circle, as shown in Fig. 4 below. We downloaded the documents produced on the
Raspberry Pi and started Green grass Core.

Edge node
Raspberry Pi,
Cloud
Hard and soft links
AWS IOT
Certificate, Pricate and
public keys

IOT nodes
Virtual machines
Instances in AWS within edge node
secure group

Fig. 3: The proposed model

 Internet of Cloud: Secure and Privacy Preserving Cloud Model with IoT Enabled Service | 97

8 Discussion and Analysis

IoT gadgets are classified as virtual machines, made into AWS, and added to the
Greengrass facility, as shown in Figure 4 below. At the time of creation, a special
announcement was made, public and private keys to all gadgets to verify them
through AWS and the Greengrass Core gadget. The connection between the two
gadgets is made through a secure system using an MQTT session called the message
vendor. Finally, Fig. 5 shows both IoT and Edge hub facilities being successfully
deployed and information trading completed at specific times.
The following are some key points to keep in mind about our AWS workspace
and model used:

– In every model, IoT gadgets are either portable or come with AWS IoT Core. 5
– In our model we added an edge view using the Green grass IoT centre view on AWS, and spoke to
it via Pi, so that we could see it as an additional middle ground between IoT and AWS IoT Core
gadgets and then a cloud.
– Each gadget needs its own authentication, private key, and CA Root announcement (this is an
AWS IoT agreement). There are different types of CA Root agreement based on IoT gadget types.
– Each gadget needs a strategy, this setting indicates which functions this gadget can perform
(interface/find/distribute/purchase, etc.)

Fig. 4: IoT-enabled nodes

98 | I. Indrani, SVB Revanth, S. Akhil Durga, Sirisha Potluri, Sachi Nandan Mohanty

Fig. 5: Successful communication and data exchange between nodes

The x-axis: days of the month, and the y-axis: the number of connections. Along
these lines, we made a gadget, strategy, and testament. At that point we joined the
arrangement to the declaration, at that point appended the authentication to the
gadget. A default strategy is demonstrated in the code listing beneath:

Listing 1: Default device policy in Amazon Web Service (AWS)

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "iot:*",
"Resource": "*"
}
]
}
 Internet of Cloud: Secure and Privacy Preserving Cloud Model with IoT Enabled Service | 99

8.1 Key points of proposed model

– Default strategy allows the gadget to play all functions (Action: iot: *) from any
remaining gadgets (App: *).
– In our model, we have developed a modified strategy to deal with an additional
layer of grass.
– Addition, activity: green grass: * means that the gadget in the green grass circle
can play all functions from and to different gadgets in the same green grass cir-
cle (App: *).

The modified version of our model appears in the following code listing. In our case,
the books are finalized using the MQTT assembly which is a machine assembly sys-
tem. MQTT is used for reasons that it is lightweight (small-sized messages and re-
quires low power), so it makes sense for forced weather (sensors as a metaphor for
real applications).

Listing 2: Modified device policy to include the edge layer in the proposed model

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"iot: Publish",
"iot: Subscribe",
"iot: Connect",
"iot: Receive"
],
"Resource": [
"*"
]
},
{
"Effect": "Allow",
"Action": [
"greengrass:*"
],
"Resource": [
"*"
]
}
]
}
100 | I. Indrani, SVB Revanth, S. Akhil Durga, Sirisha Potluri, Sachi Nandan Mohanty

In addition, note that IoT gadgets in AWS are copied as MQTT clients (if they are
visible - in what is important in our case), and MQTT clients transmit the MQTT title.
The association can be regarded as a secure channel between customers, it is
formed, at which time customers can buy from it, and various customers spread the
message to it.
Now, the Raspberry Pi setup process involves installing JAVA JDK8, green grass
files, in addition to the appropriate Core software (depending on the device used —
for us, it’s Raspberry Pi 4).
All these files have been transferred to Raspberry Pi 4 as shown in Fig. 6 below.

Fig. 6: The Raspberry Pi 4 kit setup

After the exchange of fundamental documents to the Raspberry Pi 4, we expected to

extricate them and roll out certain improvements on some arrangement records, to
coordinate the created testaments and keys. At last, we began the Green grass centre
gadget.
Fig. 7 underneath shows that our Raspberry gadget effectively filled in as an
Edge.

Fig. 7: Running Greengrass successfully on the Raspberry Pi 4 kit

 Internet of Cloud: Secure and Privacy Preserving Cloud Model with IoT Enabled Service | 101

After setting the weather and making sure it was right, we made the MQTT theme for
our situation and named it (my/point). At that time, we made the gadget to be pur-
chased on my/subject and made another gadget to be the presenter on my/theme.
All technologies can play all functions with any remaining gadgets (default strate-
gy), and all messages are successfully sold. Fig. 8 shows the different types of mes-
sages traded in one day. Meeting time is influenced by many factors including the
inertness of the organization and the pre-management stage.
The proposed IoT model has shown that we can ensure the safety and security
measures put in place before allowing the IoT enabled gadget or hub to transmit or
share its information. When properly designed and built, we ensure that our re-
sources are secure. The model shown in this paper can be used to provide secure IoT
shapes and structures with haze/edge processing layers and a combination of sen-
sors. Some real-time applications can use this model, for example, medical care,
military, debacle recovery, and many more. Allow us to look into the case of medical
services; for example, through a proposed strategic-based model, clients will be
able to trust their medical service providers to allow them to be protected to see that
they are being cared for. Medical care organizations incorporate resources into
clothing with the assurance that they will help improve employee benefits, reduce
unemployment, and reduce the cost of medical services. Another major factor in
wearable gadgets is the belief that it can provide for people with disabilities. For
example, a person with special needs will have the option of placing orders and
text, say, by moving a finger here and there. The final, but unrestricted, strategy is
the number of security customers who can work on their records. For example, indi-
viduals can classify who can see their social media posts or policies that specify the
amount of additional security is added to their record (e.g., two-factor authentica-
tion)
While the developers of IoT applications (medical services for this condition) try
to do the best for their customers, there are still a few principles that will fail. One of
the inefficiencies will be how client information is removed and how outsiders han-
dle it. It is often up to the supplier himself to ensure that they set the rules and pro-
pose an arrangement that will enable them to behave responsibly with vendors and
their customers. The same thing applies with customer segregation. More often than
not, outsiders, (for example, insurance agencies) can obtain customer data in case
they “agree” with it, and then from there, it can be dangerous to determine if it is
solid. In the following figure, on the X- axis hours of the day and on the Y- axis one
day message readings are represented.
102 | I. Indrani, SVB Revanth, S. Akhil Durga, Sirisha Potluri, Sachi Nandan Mohanty

Fig. 8: Successfully exchanged messages of different types: PublishOut.Success, Publish-

In.Success, Connect.Success, Ping.Success, Subscribe.Success

9 Conclusion
IoT devices and applications play an important role in our modern life. We can see
IoT devices almost everywhere from our homes, offices, shopping centres, schools,
airports, and many other places to provide us with secure and sought-after services.
IoT gadgets support collaboration with partners and help in understanding
business needs and outcomes. In addition, IoT-based testing and information pro-
cessing can enhance the profitability and productivity of a modern framework
In addition, IoT frameworks are implementing a variety of innovative new de-
velopment applications in a variety of areas. Retailers and many organizations are
finding a large proportion of arrangements to protect their gadgets related to mali-
cious attacks. With the large number of these gadgets associated with our private
organizations and the Internet, additional concerns and protections are also being
taken into account. We read and hear that our espresso machine is watching our
conversations; our excellent department sends our guests photos of government
agencies. Some real models emphasize the seriousness of the security vulnerabili-
ties associated with the use of IoT gadgets.
In this work, we have introduced new IoT models: standard and accessible with
insurance and security components and highlights. The proposed cloud/edge struc-
ture is maintained and tested. The low level is considered by IoT institutions devel-
oped by Amazon Web Service (AWS) as Real Machines. Medium level (Edge) has
been achieved as a Raspberry Pi 4 gear pack with a source of Greengrass Edge Envi-
ronment in AWS. The high level, which is a cloud, is eliminated using IoT cloud
 Internet of Cloud: Secure and Privacy Preserving Cloud Model with IoT Enabled Service | 103

computing in AWS. Security meetings and basic management meetings were held
between each of these layers to ensure the protection of customer data
In the future work, further research should be done on cryptographic security
strategies that are best suited to operating IoT-compressed IoT (Light Weight Crypto)
devices. It will help to ensure that clients with a variety of connections can use and
set up IoT frames without the lack of consumer integration that comes with the large
number of these IoT gadgets. In addition, there is a real need to make a variety of
information and sharing methods available for Internet-related IoT gadgets. Such
guidelines will reduce the number of unexpected vulnerabilities and related attacks
in non-identical categories.
We study the benefits and risks associated with IoT. With all the various bene-
fits, risks can be misused to harm end clients by allowing unauthorized acceptance
of sensitive information, enabling framework attacks, and endangering the health of
individuals. With IoT enabled gadgets marketed, we need to deliver them with ap-
propriate security efforts that facilitate their use, function, and join existing struc-
tures. We relied on the help of analysts to create a unique security structure to re-
duce, not really kill, security and protection opportunities, and be bright enough to
adapt to changes in the new literature development and various organizational
conditions for application.

10 References
[1] Sharma N, Shamkuwar M, Singh I, The history, present and future with IoT, Balas V, Solanki V,
Kumar R, Khari M, Internet of Things and Big Data Analytics for Smart Generation, Intelligent
Systems Reference Library, 2019, vol 154, Springer, Cham, https://doi.org/10.1007/978-3-
030-04203-5_3
[2] Khanna A, Kaur S, Internet of things (IoT), applications and challenges: a comprehensive re-
view, Wireless Pers Commun, 114, 2020, 1687–1762, https://doi.org/10.1007/s11277-020-
07446-4
[3] Chang D, Chen Y, Chen L, Chao C, Internet of things and cloud computing for future internet,
Chang S, Kim T, Peng L, Security-Enriched Urban Computing and Smart Grid, Communications
in Computer and Information Science, 2011, vol 223, Springer, Berlin, Heidelberg,
https://doi.org/10.1007/978-3-642-23948-9_1
[4] Heena W, Rajni A, Fog computing with the integration of internet of things: architecture, appli-
cations and future directions, 2018, 987-994, 10.1109/BDCloud.2018.00144
[5] Adel A, Utilizing technologies of fog computing in educational IoT systems: privacy, security,
and agility perspective, J Big Data, 2020, 7, 99, https://doi.org/10.1186/s40537-020-00372-z
[6] Anand P, Hameed P, Won-Hwa H, Hyun S, Seungmin R, Fog computing-based IoT for health
monitoring system, Journal of Sensors, 2018, vol. 2018, Article
ID 1386470, 7 pages, https://doi.org/10.1155/2018/1386470
[7] Michele D, Koen T, Nicola D, Foundations and evolution of modern computing paradigms:
cloud, IoT, edge, and fog, IEEE Access, 2019, vol. 7, pp. 150936-150948, doi:
10.1109/ACCESS.2019.2947652
104 | I. Indrani, SVB Revanth, S. Akhil Durga, Sirisha Potluri, Sachi Nandan Mohanty

[8] Aazam M, Huh E, Fog computing and smart gateway based communication for cloud of things,
2014 International Conference on Future Internet of Things and Cloud, Barcelona, 2014, pp.
464-470, doi: 10.1109/FiCloud.2014.83
[9] Lee K, Kim D, Ha D, Rajput U, Oh H, On security and privacy issues of fog computing supported
internet of things environment, 2015 6th International Conference on the Network of the Future
(NOF), Montreal, QC, 2015, pp. 1-3, doi: 10.1109/NOF.2015.7333287
[10] Chen Y, Chang Y, Chen C, Lin Y, Chen J, Chang Y, Cloud-fog computing for information-centric
internet-of-things applications, 2017 International Conference on Applied System Innovation
(ICASI), Sapporo, 2017, pp. 637-640, doi: 10.1109/ICASI.2017.7988506
[11] Elazhary H, Internet of things (IoT), mobile cloud, cloudlet, mobile IoT, IoT cloud, fog, mobile
edge, and edge emerging computing paradigms: disambiguation and research directions,
Journal of Network and Computer Applications, 2019, Volume 128, Pages 105-140, ISSN 1084-
8045
[12] Potluri S, Achyuth S, Elham Y, Mohanty S N, IOT enabled cloud based healthcare system using
fog computing: a case study, Journal of Critical Reviews, ISSN- 2394-5125, 2020, Vol 7, Issue 6,
PP: 1068-1072, doi: 10.31838/jcr.07.06.186
[13] Xuan-Qui P, Eui-Nam H, Towards task scheduling in a cloud-fog computing system, 2016 18th
Asia-Pacific Network Operations and Management Symposium (APNOMS), Kanazawa, 2016,
pp. 1-4, doi: 10.1109/APNOMS.2016.7737240
[14] Potluri S, Subbarao K, Improved quality of service-based cloud service ranking and recommen-
dation model, TELKOMNIKA Telecommunication, Computing, Electronics and Control, 2020,
Vol. 18, No. 3, pp. 1252~1258, ISSN: 1693-6930, accredited First Grade by Kemenristekdikti,
Decree No: 21/E/KPT/2018 DOI: 10.12928/TELKOMNIKA.v18i3.11915
[15] Potluri S, Subbarao K, A hybrid PSO based task selection and recommended system for cloud
data, Test Engineering and Management, 2020, Vol-83, ISSN: 0193-4120, PP: 10210 – 10217
[16] Potluri S, Subbarao K, Hybrid self-adaptive PSO and QoS based machine learning model for
cloud service data, International Journal of Control and Automation, 2020, Vol. 13, No. 2s, pp.
36 - 50, ISSN: 2005- 4297
[17] Binh M, Huynh T, Thanh B, Bao D, Evolutionary algorithms to optimize task scheduling problem
for the IoT based bag-of-tasks application in cloud–fog computing environment, Appl.
Sci, 2019, 9, 1730, https://doi.org/10.3390/app9091730
[18] Goiuri P, Iglesias-Urkia M, Barcelo M, Gomez R, Moran A, Bilbao J, Fog computing based effi-
cient IoT scheme for the industry 4.0, 2017 IEEE International Workshop of Electronics, Control,
Measurement, Signals and their Application to Mechatronics (ECMSM), Donostia-San Sebas-
tian, 2017, pp. 1-6, doi: 10.1109/ECMSM.2017.7945879
[19] Yi S, Qin Z, Li Q, Security and privacy issues of fog computing: a survey, Xu K, Zhu H, Wireless
Algorithms, Systems, and Applications, WASA 2015, Lecture Notes in Computer Science, 2015,
vol 9204. Springer, Cham, https://doi.org/10.1007/978-3-319-21837-3_67
[20] Alrawais A, Alhothaily A, Hu C, Cheng X, Fog computing for the internet of things: security and
privacy issues, IEEE Internet Computing, 2017, vol. 21, no. 2, pp. 34-42, doi:
10.1109/MIC.2017.37
Srikanth Pothuri
Marketing analytics as a Service: Secure
Cloud Based Automation Strategy
Abstract: Efficient marketing strategies for business to business marketing (content,
inbound, social media, SEO, SEM, account based, earned media, referral programs,
industry events, conversational and step ahead as resolutions) and business to
customer marketing (social network, paid media, internet, email, direct, point of
purchase, cause, conversational, earned media and storytelling) are greatly height-
ened by using cloud based marketing automation solutions. Cloud governance en-
sure to meet business desires and needs through professional practices. Significant
factors to ensure security in cloud marketing are embedded in each phase of market-
ing life cycle. Various cloud based secure marketing solutions are detailed and their
performance is compared and illustrated in this work. To meet next generation mar-
keting needs cloud based solutions are foremost preference for implementing effi-
cient marketing strategies.

Keywords: Customer Relationship Management, Cloud Marketing, Marketing

analytics as a Service, Cloud Security, Cloud Privacy

1 Introduction
Cloud based efficient solutions in marketing, sales, promotion and CRM saves mon-
ey by converting capital expenses or capital expense (CapEx) into operating expens-
es or operating expenditure (OpEx). Flexibility, availability and ease of use features
of cloud based marketing automation suite provides public cloud services based on
pay as you use model to avoid investing on servers, storage devices and other com-
puting infrastructure. For storing sensitive data, maintaining greater flexibility,
security, privacy and control, the private cloud offers necessary business solutions.
Hybrid cloud aimed at combining public and private clouds to meet the business
objectives. To bring business groups or organizations together which are having
common business interest, concern or goal in order to share CapEx and OpEx and
observe reduced costs, community cloud is an abundant choice [1-3]. Various cate-
gories of clouds which are available to meet business objectives of diverse forms of
organizations is given in fig. 1.

||
Srikanth Pothuri, Research Scholar, ICFAI Business School, ICFAI University Dehradun, Dehradun,
Uttarakhand 248197, India, [email protected]

https://doi.org/10.1515/9783110732573-006
106 | Srikanth Pothuri

Public cloud
Converts CapEx to OpEx
Provides Pay as you use model
Allows several users

Private cloud
Control existing CapEx
Supports to redce OpEx
Limited to single tenant

Hybrid cloud
Combines public and private clouds
Supports to costs by manipulating CapEx and
OpEx
Business resource portability

Community cloud
Supports to costs by manipulating CapEx and
OpEx
Business resource portability
Bring business groups or organizations together

Fig. 1: Cloud deployment models for business and management

 Marketing analytics as a Service: Secure Cloud Based Automation Strategy | 107

2 Cloud Governance through Marketing Automation:

To Meet Business Desires
Cloud platform provides expert governance solutions to meet business needs, objec-
tives and desires as shown in fig. 2. Marketing strategy includes the plan as given
below.

– Identify: Ask yourself -Who you are and what you are for 1
– SWOT: Examine yourself- What are your strengths, weaknesses, opportunities, limitations
– Target- Identify yourself- Who is your target market, their behaviour and needs
– Inspection- Examine yourself- Where your target and your contributions intersect
– USP- Create yourself- What is your unique selling point
– Refine- Review yourself- what is your feedback and corrective measures

SWOT
Identify
Examine yourself-
Ask yourself -Who
What are your
you are and what
strengths and
you are for
weaknesses

Refine
Target
Review yourself-
Identify yourself-
what is your
Who is your target
feedback and
market, their
corrective
behavior
measures

USP Inspection
Examine yourself-
Create yourself- Where your target
What is your and your
unique selling contributions
point intersect

Fig. 2: Cloud governance solutions for efficient marketing strategies

108 | Srikanth Pothuri

Cloud governance lets your organization to bring everything together to satisfy vari-
ous business needs: business monitoring, recommendations, operations manage-
ment, cost transparency, cost optimization, scheduled policies, lead dashboard,
lead analytics, report generation, auditing, budget management, and security and
compliance governance. The major disciplines cloud governance are shown in
fig. 3.

Cost Management
To evaluate and monitor

Security baseline
To ensure security

Resource consistency
To ensure resource consistency

Identity baseline
To identify roles and
assignments

Development accelaration
To standardize processes

Fig. 3: Major disciplines of cloud governance

 Marketing analytics as a Service: Secure Cloud Based Automation Strategy | 109

3 Significant Factors Ensuring Security in Cloud

Based Marketing Automation Process
The three main areas of security concerns in the cloud-based applications are ac-
cess, data, and platform [4-6]. Let us examine these in detail.

3.1 Access security

Cloud offers an integrated data storage, computing and networking mechanism.
“Who can access the data” is firmly administered by having access rights. SaaS
providers add additional layer of security and privacy to stop unauthorized access to
information in cloud.

3.2 Data security

Recent applications need infrastructure for data security to meet ongoing business
demands and needs. The increasing need of data storage, data processing and data
analytics on the cloud, has also improved the need and responsibility of data securi-
ty on cloud service providers.

3.3 Platform security

A cloud platform is a physical system that offers, houses and distributes infor-
mation. Cloud platform delivers an on-demand platform and environment for de-
signing, developing, challenging, delivering, and handling software applications.
Significance factors ensuring security in cloud based marketing automation
process at a glance is given in fig. 4.

4 Automation of Marketing Life Cycle Through Cloud

Based Secure Marketing Solution
The practical framework, RACE (Reach, Act, Convert, and Engage) can be automated
through cloud based secure marketing solutions [7-10].
The RACE Framework consists of these four steps and online marketing activi-
ties are designed to help marketing people to engage their customers throughout the
customer lifecycle. The framework is demonstrated in fig. 5.
110 | Srikanth Pothuri

Service Delivery Business terms

Service definition Insurance
Roles and responsibilities Business policies
Service management Commercial terms
Service availability Publicity
Service continuity Operational reviews

Legal protections
Data Assurance
Indemnification
Data management
Intellectual property
Data security
Limitation of liability
Ownership and usage rights
Warranties
Data conversion

Fig. 4: Significance factors ensuring security in cloud based marketing automation process

4.1 Reach
Reach includes building product awareness, media usage and visibility through mes-
saging, live chat, website visits, social media posts, advertising and mouth publicity.
It ensures maximizing the reach over time of the customer by creating several interac-
tions using different ways of paid, unpaid, owned, public and earned media networks
and touchpoints. This is to bring awareness of a particular brand or product.

4.2 Act
It is a short form of interact. With interest and intent, sales team get leads through
conversations, chat, booked calls, webinars, email marketing, list building and
 Marketing analytics as a Service: Secure Cloud Based Automation Strategy | 111

more other means of interaction. For many categories of businesses, particularly, B-

to-B, this step can bring decent quantity of leads.

Act: For interest and intent

Reach and Traffic: For awareness

Conversions
Consistent trust building content
Live chat
Lead magnets
Booked calls
Inbound SEO content
Webinars
Word of mouth referrals
Email marketing
List building

The marketing
lifecycle

Engage: For loyalty

Convert: For Decision

Maintaining relationships
Quotes
Customer reviews
Samples
Launching new content
Previous positive customer review
Launching new products
sales
Continually and consistently building trust

Fig. 5: The marketing life cycle- RACE framework

112 | Srikanth Pothuri

4.3 Convert
This step ensures online or offline lead conversion to sale. It includes getting your
target audience and customers to take the next important step of payment through
online transaction, ecommerce or offline modes. This step uses product quotes,
product samples, previous positive customer reviews and sales information to en-
gage the customer.

4.4 Engage
This is a long-term commitment and engagement, evolving a long-term association
and relationship with initial and first-time product buyers to form customer loyalty
as recurrence and repeat product purchases. Engage is ensured by using communi-
cations on your web site, social network posts and active presence, email interac-
tions, direct communications and messaging. Loyalty of the customer is maintained
through healthy relationships, competent reviews, launching new products, con-
tinuous positive feedback and consistently building trust.

5 Various Cloud Based Secure Marketing Solutions

5.1 Adzooma software
Adzooma is a digital publicity and advertising platform that supports and benefits
businesses of all dimensions of the organizations to expand marketing related oper-
ations and strategies by improving and optimizing sales promotions across Google
marketing platform, Facebook influencer marketing platform and Microsoft market-
ing and advertising center. Using the opportunity engine platform and ma-
chine/deep learning processes, experts can view programmed and automated rec-
ommendations about recommended keywords, smart targeting, budget and
economy management, CRO- conversion rate optimization and LPO- landing page
optimization.
Organizations are using Adzooma to study and analyze recommended data and
inevitably break in proceedings for ongoing promotions or proposals on low per-
forming keywords. Marketing managers can constitute automation rules by outlin-
ing precise conditions and situations to obtain reports about the budget which is
overspend across promotions. Furthermore, it lets cloud users manage several ac-
counts on a centralized environment and filter account particulars by date sort.
Marketing hierarchies can use Adzooma for customizable analysis and collect
stakeholders. Pricing is based on periodic basis (say monthly) subscriptions and
 Marketing analytics as a Service: Secure Cloud Based Automation Strategy | 113

payments are extended via feedback, live chat, email, slack, phone and other online
or accessible processes.

5.2 Bitrix24 software

Bitrix24 is a client or customer supervising solution that offers the infrastructure
and platform for businesses to establish and track relations with potential and pro-
spective customers and partners. Using this software, users maintain and manage
log of interactions to capture and pile lead data, produce sales reports and accom-
plish breakdown of target audiences. Leads which are received from a source can be
fed directly as an entry to the CRM, create message template, sending communica-
tion to leads, identify potential communication to interact with clients and schedule
meetings to meet targets. Users can produce custom-made statements for clients by
automatically maintaining the client information and sending the same to them
through email.
Bitrix24’s sales funnel module offers an overview of sales communications and
transactions. With the sales control panel/dashboard, sales managers can see the
comparison of deals, rating and sales of all team members.

5.3 WebEngage software

WebEngage is a customer relationship management platform provides marketing
automation, customer data management, user engagement and maintenance not
only for large sized business organizations but also SMBs. WebEngage sales control
panel/dashboard allows to identify potential customers, aim for better client inter-
actions, marketing analytics, fixing campaign calendar and tracking growth metrics
for great journey towards healthy targets.
This software has wide range of use across the globe for various domains such
as ecommerce, education, marketing analytics, business services, social media,
healthcare, retail, manufacturing, gaming, banking, Financial Services and Insur-
ance (BFSI) and many more.

5.4 BetterMetrics software

BetterMetrics is a resolution in marketing domain using which report automation,
next generation digital marketing using pay-per-click (PPC) strategy and integration
with various ongoing and on demand data management applications such as Google
sheets (Google analytics), Data Studio (Google interactive dashboard) and Big Query
(cloud data warehouse) solution that enables digital marketers to gather and aggre-
gate all their PPC data in a centralized location. This software is available on pay per
114 | Srikanth Pothuri

use model (subscription based- monthly). Advantages of this software are mining for
better metrics for users unified reporting and comprehensive examination.

5.5 EngageBay software

EngageBay is a combined solution for various domains such as marketing, sales,
promotion and CRM. This software mainly meant and designed for SMBs to obtain,
absorb and transform website visitors into prospective customers. This cloud-based
software allows business organizations to use this as a marketing tool to construct
healthy relationships with customers and preserve them for long time.
Advantages of EngageBay are email marketing (for communication), landing
pages (for optimization), live chat/helpdesk (for engaging), ticketing (for announc-
ing), telephony (for broadcasting), appointment scheduling (for publicizing), con-
tact management (for spreading) and more. It has eminent features namely built-in
CRM support, marketing capabilities, lead conversion policies, email templates,
social media visibility, forms and popups for better communication and automated
mails and communication.

5.6 Agile CRM software

Agile CRM is a cloud enabled CRM resolution designed for SMBs. It deals contact
supervision, telephony techniques, appointment arranging, marketing process au-
tomation, project organization, landing-page designer and knowledge and infor-
mation base functionalities.
This software also provides email services by integrating and syncing data with
Gmail, Yahoo, Microsoft Exchange, Hotmail, Office 365 and IMAP. Users can analyze
about website visitors to study customer conduct, marketing automation proce-
dures, task management and generate marketing process flow with GUI interface.
Using user friendly GUI, task management ended up in easy manner through drag-
ging, dropping, sorting, adding, and updating tasks or task status. Social media
collaboration is made feasible through integrating with Facebook, Twitter, Insta-
gram, Tumblr, WeChat, Messenger, LinkedIn and many more.

5.7 Sender software

Sender is a cloud enabled email marketing, advertising and sales solution that sup-
ports users to generate and handle email-based promotions advertisements for
SMBs. Significant features of sender software contain customizable subscription
procedures and prototypes, subscriber partitioning and management tools, email
customization and management, third-party incorporations and more.
 Marketing analytics as a Service: Secure Cloud Based Automation Strategy | 115

Users can send advertising and promotional emails, all-purpose surveys or

newssheets based on customizable prototypes. Sender also offers plugins for third-
party payment and e-commerce platforms that allow organizations to pull customer
data for emails and integrate product information to newssheets. Users can supple-
ment multimedia files such as images in various formats, audio, video and animated
and dynamic GIFs. Pricing is based on periodic basis (say monthly) subscriptions
and payments are extended via feedback through smart phone and email.

5.8 SendX software

SendX is an In-built, reasonable & customizable email marketing, advertising and
sales software for all stakeholders of the organization. It allows users to send unre-
stricted email promotions, construct email comprehensions, GUI based design of
emails with drag and drop features, email sequence, customization and automation.
This software ensures 24x7 active and live support and maintenance, better migra-
tion services, customizable and responsive email prototypes and image repository
for email promotions and campaigns. Pricing is based on periodic basis (say month-
ly and yearly) subscriptions and payments are extended via feedback through email
and live chat.

5.9 InTouch software

InTouch software is a browser (Chrome) extension service to automate LinkedIn
communications and messages. It is a safest and perfect tool for perfect for lead
generation, monitoring sales and sales management in LinkedIn. This software
prospects LinkedIn by automating the process through auto connect or join mass
scale messaging, scraper and lead board customization.
Auto connect or join mass scale messaging provides customized message man-
agement, flexible message sending and follow up. Scraper is used to define target
customers through social media network. Lead board customization provides dash-
board to start, view or manage your leads through tasks. InTouch software is used
by marketing professionals, business vendors, service provider, and recruiters.

5.10 SendinBlue
SendinBlue is a cloud enabled email based marketing, advertising and promotion
tool suited for large, medium and small size organizations. It offers automation of
marketing advertising and sales, email promotions, log of transactional emails and
customized free range SMS messages functionalities within a package.
116 | Srikanth Pothuri

SendinBlue lets users to produce, schedule, automate or deliver mobile based re-
sponsive and dynamic emails using customization tools and APIs. This software has
rich customer communication management through introducing contacts, classifying
and grouping lists, collecting and designing forms, communicative forms and many
more. SendinBlue allows reporting, content management, e-commerce administra-
tion, exporting and importing results and responsive website management.

5.11 Freshworks CRM software

Freshworks CRM is a cloud enabled efficient customer relationship management
solution that helps organizations through diverse business verticals to accomplish
their communications with prevailing and prospective customers.
Significant features consist of sales tracking, sales management, event man-
agement, single click phone and task administration. Users can also refer custom-
made majority emails from the proposed solution, and then monitor accomplish-
ments on these email activities. The inbox spontaneously highlights emails from
contacts, connections and leads that are coming up for an active response. Fresh-
works CRM trails for the dynamic and responsive web pages that prospects the con-
tacts behavior, lead board dashboard generation, task prioritization, customers and
their lead management, including discussions, touch points, activities, appoint-
ments, and responsibilities.

5.12 Routee software

Routee is a cloud-based tool intended to help business organizations of large, medi-
um and small scale. It provides process flow of message based marketing, lead gen-
eration, bulk messaging, double-factor verification, push announcements, real-time
broadcasting and data analytics.
Routee lets business organizations to send custom-made mass produced mes-
sages to the customers by including essential information. This tool enables users to
generate promotions, track communications and connections, produce subscription
practices, target and segment leads through several channels and systems.

5.13 Zoho CRM software

Zoho CRM is a cloud enabled business management and administration tool that
helps business organizations of large, medium and small scale. It deals sales, adver-
tising and marketing automation software with dashboard and helpdesk, data ana-
lytics, data visualization and customer support utilities.
 Marketing analytics as a Service: Secure Cloud Based Automation Strategy | 117

It has artificial intelligence powered sales assistant for strong customer rela-
tionship management, data indicators, dynamic statistics and well-organized inte-
gration to third parties such as Google Suite, WordPress API, MailChimp automation
tool, Evernote App and Unbounce.

5.14 Omnisend software

Omnisend is an efficient marketing, advertising and sales automation platform for
development concentrated ecommerce based business organizations. This software
permits you to enhance several networks to the same automation process flow for
unified communication via tools like email, messaging, live chat, web push notifica-
tions and announcements, and customer base management.
This can produce strong automation process flows and customize your message
and communication based on customer record, campaign arrangement and recom-
mendations through shopping behavior. Omnisend provides integration with the
popularity of ecommerce applications platforms and other marketing management
tools. Pricing is based on periodic basis (say monthly) subscriptions and payments are
extended via feedback through smart phone, live chat, knowledge sharing and email.

5.15 AiHello software

AiHello software is mainly aimed for Amazon pay-per-click (PPC) operations and
promotions. It has significant modules, auto pilot (for efficient ecommerce market-
ing, advertising and sales) tool, bidding optimization (for pay-per-click based top-
notch automation and bidding) tool, Artificial intelligence engine (for intelligent
business automation), advanced algorithms library (for restocking support and
sales forecasting), real-time analytics (for examining and analysing customer activi-
ties periodically), multichannel analytics (for integrating easily with different date
sources, warehouses and applications such as Shopify, Woocommerce, and Ma-
gento), ecommerce application (to support connection with various ecommerce
platforms) and marketing analytics (for information management).

5.16 Mailchimp software

MailChimp is a cloud and web-based email marketing, advertising and sales auto-
mation service with millions of global users. The tools lets users to share newssheets
or letters on several social networks to examine and track the customer behavior
and engagements. This tool offers customer targeted emails, run Facebook ad pro-
motions and campaigns, systematize follow-ups and observe promotion and cam-
paign activities development and progress and many more.
118 | Srikanth Pothuri

Using this tool, users can gather and evaluate their email replies. The applica-
tion offers an efficient GUI support for graphical representations and demonstra-
tions of data and customer connections/communications in various formats. Mail-
Chimp lets users to plan, share and identify email newssheets and letters from any
location all over the world. This tool provides interactive user friendly drag & drop
graphical user interface, data analytics tools, customized report generation, mobile
based application support for Android and iOS, customer relationship management,
social media activity subscription, campaign management and ecommerce based
retail marketing. Pricing is based on periodic basis (say monthly and yearly) sub-
scriptions and payments are extended via feedback through email and live chat.

5.17 HubSpot Marketing Hub software

HubSpot is a prominent cloud based advanced platform. Marketing Hub supports
companies for their web presence for lead generation, analyze return on investment
(ROI), marketing hub for business to business and business to customer service,
software utilities and libraries for marketing automation, state of the art technolo-
gies, accounting packages, construction strategies, retail business policies, real
estate approaches and more.
This software provides various solutions to invite more visitors, adapt more
leads, accomplish more deals and finally satisfy more customers. It has integrated
blogging tools, and content design tools, SEO optimization, lead conversion tech-
niques, construction of landing pages, call to action (CTA) tools, live chat and world
class state of the art marketing automation. It has a powerful hub to automate mar-
keting, sales, service, CRM and other essential organization related strategies.

5.18 Constant Contact software

Constant Contact offers a variety of in-built marketing automation applications
intended to assist small business organizations and nonprofit organizations to de-
velop their customer bases and cherish customer relationships. This tool has email
and event marketing, social promotions and events, and event reporting, sur-
vey/review/assessment management and bid management.
Constant Contact permits business organizations for customer relationship
management using spreadsheets, emails, messages, web forms, web activities, con-
tacts, social media activities and other customer related activities. Constant Contact
deals both online and offline customer engagement activities through seminars,
training, interactive programs and other campaigns.
 Marketing analytics as a Service: Secure Cloud Based Automation Strategy | 119

5.19 Wrike software

Wrike is a cloud enabled business management and administration tool that helps
business organizations of large, medium and small scale. It provisions remote work
for all marketing verticals through Gantt charts, event calendars, workload balance
sheets, efficient resource management, custom control panel and dashboards, and
real-time analytics. This software provides hierarchical structuring through files,
folders, events, projects, modules, jobs and tasks. It has custom made templates,
libraries, extensions, tools, integrations (Salesforce, LinkedIn, Dropbox, Google
suite, Slack, facebook and Adobe Creative Cloud) API for efficient marketing, adver-
tising and sales automation platform. Pricing is based on periodic basis (say month-
ly and yearly) subscriptions and payments are extended via feedback through email
and live chat.

5.20 Semrush software

SEMrush is an efficient marketing, advertising and sales automation solution aimed
to benefit organizations and enterprises streamline the processes associated to key-
word based research, search engine optimization, advertising promotions and activ-
ities, sales monitoring and more. Marketing professionals can observe improvement
in their domains through this tool with improved traffic, movement, keyword classi-
fication and ranking, keyword usage metrics, organic research tool, content man-
agement strategies.
SEMrush allows organizations and businesses to study the responsiveness of
website content across various social networking sites and accordingly release posts
across Facebook, Twitter, Instagram, Tumblr, WeChat, Messenger, LinkedIn and
many more.
On a scale of 1-5 the impressions of various marketing solutions are captured for
different parameters. Recommendation is a label from a set of labels [“not recom-
mended”, “likely recommended”, “more likely recommended”, “highly recom-
mended”]. Analysis is displayed in table 1.

Tab. 1: Comparison of various cloud based marketing solutions

S.No Software Overall Ease of use Value for Customer Functional- Recommen-
rating money support ity dation

1 Adzooma 4.18 4.5 4.5 4.5 4.0 Likely rec-

ommended
2 Bitrix24 4.0 4.0 5.0 3.0 4.0 Likely rec-
Software ommended
120 | Srikanth Pothuri

S.No Software Overall Ease of use Value for Customer Functional- Recommen-
rating money support ity dation

3 WebEngage 4.0 4.0 5.0 4.0 4.0 Likely rec-

ommended
4 BetterMet- 4.2 5.0 5.0 4.5 4.0 More likely
rics recom-
mended
5 EngageBay 4.63 4.5 5.0 5.0 4.5 Highly rec-
ommended
6 Agile CRM 4.12 4.0 5.0 3.5 5.0 Likely rec-
ommended
7 Sender 4.6 5.0 5.0 5.0 4.5 Highly rec-
ommended
8 SendX 4.74 4.5 4.5 4.5 4.5 Highly rec-
ommended
9 InTouch 4.88 5.0 5.0 5.0 5.0 Highly rec-
ommended
10 SendinBlue 4.59 4.5 4.5 4.5 4.5 Highly rec-
ommended
11 Freshworks 4.62 4.5 4.5 4.5 4.5 Highly rec-
CRM ommended
12 Routee 4.3 4.0 4.0 4.5 4.0 More likely
recom-
mended
13 Zoho CRM 4.2 4.0 4.0 4.5 4.5 More likely
recom-
mended
14 Omnisend 4.77 4.5 4.5 4.5 4.5 Highly rec-
ommended
15 AiHello 4.6 4.0 5.0 5.0 5.0 Highly rec-
ommended
16 MailChimp 4.48 4.5 4.5 4.0 5.0 More likely
recom-
mended
17 HubSpot 4.49 4.5 4.0 4.5 4.5 More likely
Marketing recom-
Hub mended
18 Constant 4.0 4.0 4.0 4.0 4.0 Likely rec-
Contact ommended
19 Wrike 4.28 4.0 4.0 4.5 4.0 More likely
recom-
mended
20 SEMrush 4.65 4.5 4.5 4.8 4.9 Highly rec-
ommended
 Marketing analytics as a Service: Secure Cloud Based Automation Strategy | 121

Comparison of various cloud based marketing solutions are shown in fig. 6.

SEMrush
Wrike
Constant Contact
HubSpot Marketing Hub
MailChimp
AiHello
Omnisend
Zoho CRM
Routee
Freshworks CRM
SendinBlue
InTouch
SendX
Sender
Agile CRM
EngageBay
BetterMetrics
WebEngage
Bitrix24 Software
Adzooma

0 1 2 3 4 5

Recommendation Functionality Customer support

Value for money Ease of use Overall rating

Fig. 6: Analysis of various cloud based secure marketing solutions

Comparison of the given cloud based marketing models on their overall rating and
recommendation label is given in fig. 7.

6 Results and Discussion

All these cloud solutions for best practices in marketing are revealing the following.
– Stand out by accepting, understanding and appreciating your customers
– Improve the focus and concentration of your marketing team to look up the full
context and perspective of customer requirements and demands
122 | Srikanth Pothuri

– Engage and participate with seamless, continuous and personalised experienc-

es and understandings
– Automate RACE framework for reach, act, convert and engage
– Ensure security across marketing activities through secure cloud based market-
ing software/tools
– Guarantee cloud governance in marketing verticals to observe efficient cost
management, resource usage and development
– Build customer expectation and trust
– Extend customer relationships and associations with greater perception and
agility
– Accomplish your assurances to customers and the product/brand
– Use real-time business computation and intelligence to empower fast data anal-
ysis
– Use cloud based marketing automation to assure dynamic and flexible decision-
making across the organisation
– Delight customers through optimising marketing presentation
– Observe demand and development for B2B marketing with cloud based market-
ing automation tools
– Empower marketers to productively generate potential lead conversions, and,
eventually, drive more business
– Create a business test case to promote your marketing practices and strategies
– Apply “customer first” strategy
– Obtain efficient cloud based solution to create modern multichannel marketing
– Perceive efficient and easy marketing management with efficient SEO tools built
on real-time data
– Complete and in-built marketing segmentation functions and verticals
– Custom-made communication
– Real-time data analytics on customers data for quick and effective decision
making
– Dynamic generation of customer profiles across all networks can be analyzed
identify their behavior and requirements
– Lead Management functions integrated to marketing and sales departments for
cross-departmental analysis
– Identify organization needs and can deploy a cloud type: public, private, hybrid
or community
– Identify marketing needs and requirements and get the service as marketing
analytics as a service form the cloud service provider
– Identify integration of various departments in an organization and same can be
perceived by using cloud based marketing software
– Enable personal, associated journeys that will increase customer conversion
– Drive loyalty and escalate marketing’s contribution and involvement to the
bottom line
 Marketing analytics as a Service: Secure Cloud Based Automation Strategy | 123

100%

80%

60%

40%

20%

0%
Overall rating Recommendation

Adzooma Bitrix24 Software

WebEngage BetterMetrics
EngageBay Agile CRM
Sender SendX
InTouch SendinBlue
Freshworks CRM Routee
Zoho CRM Omnisend
AiHello MailChimp
HubSpot Marketing Hub Constant Contact
Wrike SEMrush

Fig. 7: Comparison of the given cloud based marketing models based on their over-all rating and
recommendation

7 Conclusion
There are various other cloud based marketing tools such as Sap Marketing Cloud,
Lengow, Basis, ZoomInfo, Evocalize, Jilt, FIREBusinessPlatform, Promoboxx, Ad-
sale, Influx MD, MailOptin, Involve.me, FunnelMaker, Converttri, Privy, Oracle
Datamanagement Platform, SutiCRM, Automizy, Morph.ai, SlickText, Vitalblocks
CRM, Synerise, Goody, MoonMail, EmailDelivery.com, Tabrasa, TractionNext,
Whatagraph, Texting Base, TargetEveryone, SugarCRM, SendPulse, SalesSeek,
Skyword, Rejoiner, Pure360, Prisync, QZZR, Post Planner, Voogy, Mission Suite,
iSalesCRM, marketing Automation Professional, Marketing Automation Enterprise,
BOTNATION AI, Mailify, LeadByte, Blackbaud Luminate Online, Instapage, Inter-
com, Greenrope, Genoo, Higher Logic, FreshMail, EverString, Ecrion, Dialog Insight,
AutoManager, ConvergeHub, Contactually, Click2Mail, ConnectAndSell, ClickDi-
124 | Srikanth Pothuri

mensions, Brightpod, C2CRM, Bronto Marketing Platform, BigMarker, BannerFlow,

PromoNavi, Salesgenie, Bonzo, Nimbus Omnichannel, Insightly Marketing,
BounceX, INBOX25 and many more. Marketing life cycle yields more production
with using these efficient cloud based tool by implementing best practices.

8 References
[1] Kirtiş K, Karahan F, To be or not to be in social media arena as the most cost-efficient market-
ing strategy after the global recession, Procedia - Social and Behavioral Sciences, 2011, Vol-
ume 24, Pages 260-268, ISSN 1877-0428, https://doi.org/10.1016/j.sbspro.2011.09.083
[2] Muliana M, Nek K, Azhari A, An overview of private preschool in malaysia: marketing strategies
and challenges, Procedia - Social and Behavioral Sciences, 2014, Volume 130, Pages 105-113,
ISSN 1877-0428, https://doi.org/10.1016/j.sbspro.2014.04.013
[3] Joseph J, Jeffery S, Mark R, Ramirez E, Martinez J, Green marketing strategies: an examination
of stakeholders and the opportunities they present, J. of the Acad. Mark. Sci, 2011, 39, 158–
174, https://doi.org/10.1007/s11747-010-0227-0
[4] Nancy W, Innovative marketing strategies in academic libraries: an overview, innovations in
the designing and marketing of information services, Jeyasekar J, Saravanan P, IGI Global,
2020, pp. 1-15. http://doi:10.4018/978-1-7998-1482-5.ch001
[5] John P, Steven M, Marketing strategies that make entrepreneurial firms recession-resistant,
Journal of Business Venturing, 1997, Volume 12, Issue 4, Pages 301-314, ISSN 0883-9026,
https://doi.org/10.1016/S0883-9026(97)89449-9
[6] Baena V, Online and mobile marketing strategies as drivers of brand love in sports teams:
findings from real madrid, International Journal of Sports Marketing and Sponsorship, 2016,
Vol. 17, No. 3, pp. 202-218, https://doi.org/10.1108/IJSMS-08-2016-015
[7] Aspelund A, Madsen K, Moen, A review of the foundation, international marketing strategies,
and performance of international new ventures, European Journal of Marketing, 2007, Vol. 41
No. 11/12, pp. 1423-1448, https://doi.org/10.1108/03090560710821242
[8] Gabrielsson P, Gabrielsson M, Seppälä T, Marketing strategies for foreign expansion of com-
panies originating in small and open economies: the consequences of strategic fit and perfor-
mance, Journal of International Marketing, 2012, 20(2), 25–48,
https://doi.org/10.1509/jim.11.0068
[9] Cacciolatti L, Hee S, Revisiting the relationship between marketing capabilities and firm per-
formance: the moderating role of market orientation, marketing strategy and organisational
power, Journal of Business Research, 2016, Volume 69, Issue 12, Pages 5597-5610, ISSN 0148-
2963, https://doi.org/10.1016/j.jbusres.2016.03.067
[10] Adjei M, Clark M, Relationship marketing in a B2C context: the moderating role of personality
traits, Journal of Retailing and Consumer Services, 2010, Volume 17, Issue 1, Pages 73-79, ISSN
0969-6989, https://doi.org/10.1016/j.jretconser.2009.10.001
Sachi Nandan Mohanty, Gouse Baig Mohammad, Sirisha Potluri,
P. Ramya, P. Lavanya
Next Generation Cloud Security: State of the
Art Machine Learning Model
Abstract: Recent days have seen an obvious shift in computing platforms and envi-
ronments with the arrival of cloud computing. Most of the organizations moving
towards cloud technology to satisfy varied user demands and requests, and to guard
and safeguard the transactions and operations of the organizations. Over the cloud
platforms, it is extremely essential to produce a secure and strong environmental
support to ensure security and privacy. Machine learning (ML) based models have
well-tried their significance to anticipate in finest outcomes to boost the choice of
cloud based long-term course of actions. Efficient machine learning models have
long been utilized in several application domains that required the identification
and prioritization of adverse factors for a threat. The main objective of this paper is
to propose a best strategic cloud security enhancement model for next generation
computing standards. Efficient machine learning algorithms like convolution neural
network gives automatic and responsive approaches to reinforce security in a cloud
environment. These models give solutions that incorporate holistic approaches for
secure enterprise knowledge throughout all the cloud applications.

Keywords: Machine Learning, Cloud Security, Cloud Privacy, Cloud Computing

||
Sachi Nandan Mohanty, Department of Computer Engineering, College of Engineering Pune,
Wellesley Rd, Shivajinagar, Pune, Maharashtra 411005, India, [email protected]
Gouse Baig Mohammad, Department of CSE, Vardhaman College of Engineering, Hyderabad,
Telangana 501218, India, [email protected]
Sirisha Potluri, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana
501203, India, [email protected]
Ramya Reddy Padala, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana
501203, India, [email protected]
Lavanya Reddy Padala, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana
501203, India, [email protected]

https://doi.org/10.1515/9783110732573-007
126 | Sachi Nandan Mohanty, Gouse Baig Mohammad, Sirisha Potluri, P. Ramya, P. Lavanya

1 Introduction
1.1 Cloud computing
Cloud computing is a new distributed and computing technology using which we
can achieve enhancement in the use of virtualized resources. These virtualized re-
sources are designed for the service provisioning of the end users in the dynamic
cloud environment in order to provide reliable and trusted services to cloud users all
the time. The main advantage of cloud computing environment is that we can im-
prove and optimize the utilization of physical resources as much as possible. The
beauty of this technology mainly lies by applying abstraction and isolation of physi-
cal resources so as to reduce the requirement of the hardware equipment. To handle
the dynamic tasks which are highly based on the demand, the cloud computing is
making use of virtualization. Virtualization is the technology using which many
virtual machines for a given single physical resource/server can be created to han-
dle/execute the multiple tasks. Resource management plays an important and vital
role here because of this dynamic provisioning of the resources to handle the on
demand service requirement of the cloud computing environment. The main objec-
tives of the resource management are to maximize the utilization time of resources,
to reduce the task execution time, to achieve high system throughput and to im-
prove load balancing. In any cloud system architecture mainly there are three main
modules namely application (source for task submission), mapping algorithms (us-
es various algorithms to map the submitted tasks to physical resources) and virtual
machine. Using mapping algorithms a set of cloudlets are mapped to resources by
estimating the expected time of completion of cloudlets and then execute the cloud-
lets on selected resources [1-2].

1.2 Machine learning

Machine learning based systems were proved as best solutions to every alternative
in far and wide. Foremost cloud service providers like Amazon internet services,
Microsoft azure, and Google cloud implements machine learning for their enterprise
level security. All major cloud service providers use totally different machine learn-
ing systems which are different from one another in terms of their functionality and
hierarchy [3].

1.3 Advantages of machine learning systems on the cloud

The advantages of the machine learning systems on the cloud are summarized as
below [4-5].
 Next Generation Cloud Security: State of the Art Machine Learning Model | 127

– Machine learning based cloud systems are price competent solutions 1

– Integration of machine learning models with computer program Interface(s) (APIs) and computer
code developer kits (SDKs) is effortlessly well-suited
– Machine learning based cloud secure solutions has intelligent future and able demand
– Scalability is enhanced in pay per use model of cloud with machine learning based applications
– Machine learning based cloud applications provides enhanced cloud based solutions

1.4 Limitations of machine learning systems on cloud

The limitations of the machine learning systems on the cloud are summarized as
below [6].

– Lack of specialized skill personnel and expertise due to insufficiency and absence of domain 1
knowledge
– Limited support to accomplish information integration and attachment at enterprise level
– Increased deployment cost to achieve required computational power through infrastructure and
workforce
– Basic engineering, easy and medium complex problems are addressed by using open source
machine learning tools such as Tensor Flow, MXNet and CNTK. Whereas these models may not
give efficient solutions for the problems with greater complexity and convolution

1.5 Literature survey

– B. Nassif et al. stated that the best way to ensure security in cloud is by using
machine learning models. This paper summarizes about different types of cloud
security threats, various machine learning techniques and performance related
outcomes. According to the survey results, they stated that support vector ma-
chine model is one of the best machine learning technique, true positive rate
and training time are the most applied and least applied metrics respectively,
KDD and KDD CUP’99 are mostly used datasets [7].
– Han Qiu et al. stated that IoT based remote healthcare system uses cloud based
data analytics to build body sensor network. To make certain security and pri-
vacy preservation in body sensor network cloud computing model is using se-
lective encryption techniques through machine learning approach. Light weight
selective encryption technique design is based on the efficient classification
machine learning models [8].
– Sirisha Potluri et al. stated that detection and prevention mechanisms related to
DDoS attacks in cloud necessitate efficient dynamic usage of the cloud re-
sources. To handle these adverse effects, cloud uses various defence mecha-
nisms using machine learning [9].
128 | Sachi Nandan Mohanty, Gouse Baig Mohammad, Sirisha Potluri, P. Ramya, P. Lavanya

– H. Karimipour et al. stated that security threats and vulnerabilities in digital

communication technology are addressed by using machine learning tech-
niques for efficient and reliable power distribution. Symbolic dynamic filtering
technique is used to reduce computational encumber in smart grids while dis-
covering laid-back interactions among the nodes [10].
– Wei Wu et al. stated that unsupervised machine learning model is used for Bot
detection during traffic analysis based on similarity measures. An efficient clus-
tering algorithm is used to identify majority clusters with most flows, elimina-
tion of duplicate flows, and computation of similarity analysis [11].
– Ünal Çavuşoğlu et al. stated that hybrid and layered intrusion detection system
uses a combination of various machine learning and feature selection techniques
to provide high performance intrusion detection in different attack types [12].
– Sirisha Potluri et al. stated that fog computing network is designed for distributed
and decentralized computing platform by using IoT infrastructure to send and re-
ceive the data. Various applications of fog computing includes e-commerce, mo-
bile, agriculture, auto mobiles, smart grids, healthcare and data analytics [13].
– Mahmud et al. stated that fog-based IoT healthcare solution structure integrates
cloud and fog based services in interoperable healthcare solutions. Using iFog-
Sim simulator the results are analysed in view of distributed computing to show
latency reduction, communication cost optimization and efficient energy con-
sumption [14].
– Subramanian, E. et al. stated that with the aim of security and privacy to safe-
guard the cloud transactions, it is extremely necessary to provide a secure, safe
and robust machine learning based solution across cloud vicinity. Various ma-
chine learning approaches such as linear regression, convolution neural net-
work, and support vector machine provide enhanced cloud based solutions for
ensuring security and privacy [15].
– Mbarek Marwan et al. stated that using support vector machines and fuzzy C-
means clustering classification of image pixels can be done more efficiently. Us-
ing conventional two layered architecture, simultaneous image segmentation
and data protection is achieved in medical image data [16].

2 Machine Learning Based Cloud Solutions to

Address Research Issues and Challenges
2.1 Privacy and security aware safety measures
Cloud computing environment make use of large amounts of datasets for storage
and processing. On premise data storage, handling, processing is completely risk
free. But if the data is getting saved in other places there is a greater risk involved in
 Next Generation Cloud Security: State of the Art Machine Learning Model | 129

that. Cloud service providers, intermediaries or the companies involved in hosting

of cloud platforms may disclose the data which is being in access to them. Due to
which data may be tampered or lost. In either of the cases it leads to security and
privacy issues. Due to these reasons of security and privacy, cloud computing is
taking slightly a back step apart from its wider adoption. Machine learning based
cloud computing models and approaches gives next generation privacy and security
in computing platforms [17].

2.2 Performance and its measurement

In cloud environment, this issue is very much important as it is one of the measur-
ing techniques in computing environment. Performance is a major concern in cloud
system which affects the delivery of various services, revenue generated due to ser-
vice delivery and the number of customers involved in the process [18].

2.3 Attentive solutions through reliability and availability

Reliability defines the certain expected functioning of the system under selected
time interval and with given conditions. Availability of any system can be defined
as degree in which the set of resources are available and these available resources
are used whenever and wherever there is a need. These two factors comes together
and defines the strength and potency of the system [19].

2.4 Accustomed solutions through scalability and elasticity

Scalability can be defined as its ability to adapt to the changes. When there is a
need of resources by various cloud users, the system should take care of workloads
efficiently. Elasticity of the system can be defined as the factor up to which it can
handle the workloads efficiently. Elasticity ensures allocation and deallocation of
resources to handle the needs of on demand users [20].

2.5 Adaptive solution through interoperability and portability

Interoperability of cloud computing environment can be defined as the power or
ability which is adopted by the system to run various services which are from vari-
ous vendors. Its associated factor interoperability can be defined as the way in
which different platforms are effectively used among different cloud service provid-
ers. Portability is an important factor which defines the ability to transfer the related
130 | Sachi Nandan Mohanty, Gouse Baig Mohammad, Sirisha Potluri, P. Ramya, P. Lavanya

data applications from one CSP to another CSP without much dependency involved
in it [21].

2.6 QoS driven resource management and scheduling

Resource management can be defined as a factor which represents the efficient
usage and maintenance of the resources. Various types of resources are mapped to
the submitted jobs so as to schedule them in a more efficient manner. A better re-
source scheduling or provision can be achieved by using a better scheduling policy.
By using these scheduling algorithms, we can optimize cloud services to achieve
high level of quality of service. There are many categories of users, tasks and re-
sources using which task mapping and execution is achieved [22-24].

2.7 Efficient energy consumption infrastructure

Any cloud data centre or storage consists of infrastructure in large and voluminous
amounts. Cloud computing environment is delivering various types of services to
enormous users by using its massive infrastructure. Cooling infrastructure is needed
to compensate the heat which is produced by these servers. This Expensive infra-
structure should be developed; operated and maintained very carefully as it pro-
duces greenhouse gases which are adverse to environmental balance. Proper cool-
ing system and mechanisms are maintained by cloud centres to maintain the cloud
infrastructure as environment friendly [25].

2.8 Able distributed computing with virtualization

In distributed and cloud computing environment, the concept of virtualization plays
a very important role. This refers to the process of creating virtual machines for the
various physical resources. This concept is closely related to task scheduling pro-
cess where task queue is mapped to various virtual machines for scheduling. The
main advantages of virtualization are cost reduction, highly scalable and provide
elastic behaviour to the cloud computing environment [26].

2.9 Cost sensitive infrastructure with better throughput

According to cloud computing, bandwidth refers to the rate in which bits are trans-
ferred. This measurement is mainly useful when we are calculating throughput of
the system. Since cloud computing environment is providing high speed transfer of
data among systems it requires high bandwidth [27].
 Next Generation Cloud Security: State of the Art Machine Learning Model | 131

3 Machine Learning Based Cloud Models- Google

Cloud Machine Learning Engine: A Case Study
Various machine learning algorithms which are used in cloud security and privacy
are shown in fig. 1.

Machine learning
algorithms

Unsupervised learning Reinforcement

Supervised learning
-Clustering learning
-Classification
-Dimensionality -Learning tasks
-Regression
reduction -Skill aquisition

Fig. 1: Machine learning algorithms- Overview

In cloud computing and other environments, there are many similarities and differ-
ences in resource planning. The objective of scheduling is to achieve remarkable
difference in resource usage. In traditional and conventional resource mapping
scheduling, objects used are threads, jobs and tasks that pertain to the well-defined
well- grained mapping and scheduling of entity resources. The cloud scheduling
objects such as tasks and jobs are virtual machines (VM) belonging to the well-
grained mapping and scheduling. Google cloud is machine learning models to ob-
serve revolutionary changes in cloud resource management.

3.1 Google cloud platform introduction

There are revolutionary changes arising in hardware and software that are equaliz-
ing machine learning (ML). Google Cloud Platform contains a separate kind of prod-
uct or tools for various kinds of users such as beginners and specialists. For several
years, Machine learning has been a supreme strength of Google’s internal systems.
Google cloud platform is providing proper frameworks, techniques, infrastructure,
and information based on our requirement for data-driven systems on an outsized
scale to satisfy customer’s needs. Google is providing managed services that help
 132 | Sachi Nandan Mohanty, Gouse Baig Mohammad, Sirisha Potluri, P. Ramya, P. Lavanya

developers and scientists to make use of machine learning models. It is called

Google Cloud Machine Learning (ML) Engine. Google Cloud Machine learning en-
gines consists of advanced approaches and techniques [28]. There are many stages
in machine learning workflow and they are illustrated as below.

1 – Collect all the sources and prepare the desired data

– Code your desired model
– Train, evaluate, assess and tune your model
– Deploy your designed trained model
– Obtain Predictions
– Monitor the systematic predictions
– Manage your designed model and its successive versions

Training of machine learning model on your data ensures training of the machine
learning model, evaluating the accuracy of the model and tuning the hyper parame-
ters of the model. Subsequent to training, deploying of your model takes place. Lat-
er than for prediction requests, model uses online or batch prediction [29-32]. Moni-
toring systematic or ongoing predictions is a continuous process to evaluate the
performance of the machine learning model. The designed model has to be man-
aged for its successive versions or deliveries.
Machine learning is a data analytics approach that trains computers to do as
expected naturally similar to humans and natures: learn and get trained from the
experience. Machine learning approaches use computational techniques to “learn”
information directly from data without depend on a pre-programmed rules.
With the intensification in big data, machine learning has turn out to be a key
method for solving difficulties in areas, such as:
– Computational finance
– Image processing
– Computer vision
– Computational biology
– Energy production
– Automotive
– Aerospace
– Manufacturing
– Predictive maintenance
– Natural language processing

Machine learning approach can be illustrated as shown in fig. 2.

 Next Generation Cloud Security: State of the Art Machine Learning Model | 133

Machine
Building
learning
logical model
algorithm
Input past
Output
data Based on
Based on
learning
training
works for new
learns from
data
data

Fig. 2: Machine learning process- Outline

3.2 Various types of services

Various types of services are illustrated as shown in fig. 3.

Training service

Prediction service

Deep learning VM Various types of

image services

Notebooks

Data labelling
service

Fig. 3: Various types of services offered by machine learning algorithms

134 | Sachi Nandan Mohanty, Gouse Baig Mohammad, Sirisha Potluri, P. Ramya, P. Lavanya

Machine learning algorithms identify natural patterns in the given data that create
insight and aid you make improved decisions, forecasts and predictions. They are
used every day to make critical decisions in medical analysis, stock trade-off, energy
load prediction, and more.

3.2.1 Training service

Training service allows us to train the models and prototypes using a wide collec-
tion of different customization choices. Selection of different machine types allows
us to power your training jobs, facilitate distributed training, practise using hyper
parameter tuning, and accelerate with graphics processing unit or tensor processing
unit.

3.2.2 Prediction service

Platform prediction service lets us to offer predictions based on a trained model,

whether or not the model or prototype was trained on artificial intelligence platform.

3.2.3 Notebooks

Notebooks service allows us to create, manage and destroy virtual machine instanc-
es using virtual machine manager that are available in JupyterLab. These services
contains pre-installed packages such as deep learning, TensorFlow and PyTorch.
These instances are protected by google cloud authentication and they are easily
integrate with GitHub repository.

3.2.4 Data labelling service

Data labelling service allows us to human labelling of dataset with text, image or
video for training designed machine learning model. These services provides all the
possible labels for the given dataset and instructions in what way these labels are
given.

3.2.5 Deep learning VM image

Deep learning VM image allows us to choose from a set from set of images optimized
for data science and ML tasks.
 Next Generation Cloud Security: State of the Art Machine Learning Model | 135

3.3 How to use google cloud machine learning?

Follow the below-mentioned steps to use Google Cloud Machine Learning

– Register to your Google account with your existing credentials 5

– Produce a Google Cloud Platform Project
– Change the Cloud Machine Learning Engine and cipher Engine arthropod genus
– Congratulations, you’re exploiting Google Cloud Machine Learning Engine

3.4 Tools to interact with artificial intelligence platform

Follow the below-mentioned tools to interact with artificial intelligence platform.

3.4.1 Google cloud console

Using cloud console, we can manage and deploy loud models, versions and proto-
types. This tool acts like an interface to work with AI resources by using tools such
as cloud logging and cloud monitoring.

3.4.2 Gcloud command-line

Using Gcloud command line gcloud ai-platform, we can manage AI models, proto-
types and versions to accomplish AI tasks at the command line. Gcloud commands
and REST API are used for online predictions in AI platform.

3.4.3 REST API

REST API offers RESTful services for aiding online predictions.

3.5 Advantages of machine learning based google services

– Better productivity with fast access to invention and innovation: On a periodic
basis Goggle AI system delivers updates in an efficient manner.
– Fewer disturbances when users adopt new features and functionality: Google
gives new features and functionalities in a continuous stream.
136 | Sachi Nandan Mohanty, Gouse Baig Mohammad, Sirisha Potluri, P. Ramya, P. Lavanya

– Global work culture in employees: Employees can access the services and in-
formation from anywhere in the world through web based applications of
Google.
– Supports quick collaboration: Cloud users quickly access the data which is
available in the cloud due to high availability of the services.
– Enhanced security with skilled personnel: Google has and hires skilled person-
nel to ensure next generation security in cloud.
– Ratio of data stored in cloud and local computers: Fewer data get stored in local
devices to avoid vulnerability, whereas high amount of information and ser-
vices are made available through cloud platform.
– High availability and reliability of the services: Due to redundant data centres,
the data and services are made available to the cloud users with high availabil-
ity and reliability.
– On demand scalable access and control to users: Through powerful Google apps
users can have on demand scalable access and control over the services.
– Less expenditure and more gains: Google cloud based services allows users to
access to them with less expenditure.

3.6 Statistical facts about machine learning

– International data corporation- IDC says that overall revenue in various indus-
tries including software, hardware and other service oriented is expected more
than $156 billion in 2020. It has observed clearly with an increase of more than
12% over 2019.
– IDC’s Worldwide Semiannual Artificial Intelligence Tracker says that global
revenue will improve on by $300 billion by 2024, with CAGR (compound annual
growth rate) of more than 17%.
– Forbes- Global media magazine says that AI based pricing and promotion have
the potential to bring more than $259 billion in global market value.
– In telecom industry, explosion of data may increase the risk of security threats.
Google cloud says that data traffic rate may cross 77 exabytes per month across
the global market with a CAGR of 46% by 2022.
– Google says <0.1% of average Gmail mails is spam and in that 0.05% are useful
mails. Spam mail prediction and detection is possible with machine learning al-
gorithms.
– Intelligence through chatbots will reach $142 billion by 2024 from just $2.8 bil-
lion in 2019.
– Online education market may catch up to $350 billion by 2025 with machine
learning technology.
 Next Generation Cloud Security: State of the Art Machine Learning Model | 137

4 Commercial Web Based Clouds based on Machine

Learning
Cloud SIEM- Security information and event management system is shown in fig. 4.

Collect
Security data globally
from enterprise cloud

Cloud SIEM-
Respond rapidly and Security Detect threats using
initiate automatic information and machine learning
protection event approach
management

Investigate security
threats guided by
machine learning
appraoch

Fig. 4: Cloud SIEM- Security information and event management system

4.1 Cloud shell

Cloud shell is a web based environment which supports and manages resources by
using online terminal with the aid of utilities, command line tools and libraries.
Using cloud shell we can create cloud based applications. Google cloud shell in-
cludes cloud speech to text, app engine, data transfer and cloud build.
138 | Sachi Nandan Mohanty, Gouse Baig Mohammad, Sirisha Potluri, P. Ramya, P. Lavanya

4.2 Various tools available in cloud shell

4.2.1 Persistent disk storage

Google cloud storage offers persistent disk storage using virtual machines and vir-
tual machine manager interface on a pay per usage basis for every user. $HOME
directory is mounted to provide this storage with various files such as home directo-
ry, software, scripts and configuration files.

4.2.2 Authorization

Google cloud API makes use of command line tools with authorization using Au-
thorize Cloud Shell’s dialog.

4.2.3 Pre-configured environment variables

When the cloud shell is initiated the project will be given to gcloud
configuration system for immediate use. By using environment variable
GOOGLE_CLOUD_PROJECT application default credentials of the library and active
project also will be set in console.

4.2.4 Zone selection

Cloud shell is distributed in worldwide through various google cloud platforms.

Initially it is assigned with the closest geographical region. Later, based on the
workload it will try to migrate to other VMs.

4.2.5 Image rollout

Latest images of cloud shell gets updated with loud software development kit,
Docker and required runtime utilities and libraries.

4.2.6 Root user

When you create user account you have all the privileges on the allocated virtual
machine and you can execute workloads on it.
 Next Generation Cloud Security: State of the Art Machine Learning Model | 139

4.2.7 Available tools

These available tools are pre-installed and additional tools can be installed on a
requirement basis. Additional tools which are installed on the virtual machine will
not persist after instance gets terminated.

4.2.8 Language support

Various languages namely Java, Go, Python, Node.js, Ruby, PHP and .Net provides
pre-installed language support.

4.2.9 Safe mode

Google cloud is using cloudshellsafemode=true in the URL to ensure safe mode and
fix security issues in the files

4.3 Amazon Web Services

Amazon web services –AWS is a web based cloud platform launched in 2006 and
provides various products such as Amazon sage maker, Amazon augmented AI,
Amazon forecast, Amazon translate, Amazon personalize, AWS deep learning AMI’s
and Amazon polly.

4.3.1 Microsoft Azure

Microsoft azure is a web based cloud computing platform launched in 2010 and
provides various products such as Microsoft azure cognitive service, Microsoft azure
data bricks, Microsoft azure bot service, Microsoft azure cognitive search and Mi-
crosoft azure machine learning.

4.4 IBM Cloud

IBM cloud is a cloud computing platform provides several types cloud services such
as public, private and hybrid with various products namely IBM Watson studio, IBM
Watson speech-to-text, IBM Watson text-to-speech, IBM Watson natural language
understanding, IBM Watson visual recognition and IBM Watson assistant.
140 | Sachi Nandan Mohanty, Gouse Baig Mohammad, Sirisha Potluri, P. Ramya, P. Lavanya

5 State of the art Machine learning algorithms for

cloud security
Various state of the machine learning algorithms are listed in the table 1.

Tab. 1: 6.State of the art machine learning algorithms

Algorithm Machine learning algorithm Remarks

A Review of Machine Learning Supervised, unsupervised, Security threats, issues, and

Algorithms for Cloud Compu- semi supervised and rein- associated solutions of cloud
ting Security forcement computing are stated
Assessment of machine learn- Supervised Powerful analytical methods that
ing algorithms in cloud compu- allow machines to recognize
ting frameworks patterns and facilitate human
learning are stated
Network based IDS for cloud Deep Neural Network (DNN) Reduced execution time, conver-
environment using combination based anomaly Network IDS gence time and save processing
of machine learning algorithms using a hybrid optimization power
framework (IGASAA) based on
Improved Genetic Algorithm
(IGA) and Simulated Annealing
Algorithm (SAA)
Feasibility of Supervised Ma- Supervised Demonstrates the effectiveness
chine Learning for Cloud Secu- of these models across multiple
rity datasets by improving cloud
security
Evaluating machine learning Supervised and unsupervised Detects abnormal behaviour of
algorithms for anomaly detec- services and hosts by analysing
tion in clouds metrics collected from all layers
and components of the cloud
infrastructure

Code for machine learning algorithms such as linear regression, support vector
machine and Naive Bayes classification is given as below.

Listing 1: Code for machine learning algorithms such as linear regression, support vector machine
and Naive Bayes classification

#Linear regression
lin_regr = linear_model.LinearRegression()
lin_regr.fit(X_train, y_train)
print(lin_regr.coef_)
 Next Generation Cloud Security: State of the Art Machine Learning Model | 141

print(lin_regr.score(X_test, y_test))
plt.style.use('fivethirtyeight')
plt.scatter(lin_regr.predict(X_train), lin_regr.predict(X_train) -
y_train, color = "yellow", s = 10, label = 'Training data')
plt.scatter(lin_regr.predict(X_test), lin_regr.predict(X_test) - y_test,
color = "pink", s = 10, label = 'Testing data')
plt.hlines(y = 0, xmin = 0, xmax = 100, linewidth = 5)
plt.legend(loc = 'upper right')
plt.title("Demo")
plt.show()
#Support vector machine
from sklearn import svm as s
from sklearn import metrics as m
s_class = s.SVC(kernel='linear')
s_class.fit(X_train, y_train)
y_pred = s_class.predict(X_test)
print(m.accuracy_score(y_test, y_pred))
print(m.precision_score(y_test, y_pred))
print(m.recall_score(y_test, y_pred))
#Naive Bayes Classification
import numpy as n
import pandas as p
from sklearn.naive_bayes import GaussianNB
from sklearn.metrics import confusion_matrix, accuracy_score
NB_class = GaussianNB()
NB_class.fit(X_train, y_train)
y_pred = NB_class.predict(X_test)
ac_sc = accuracy_score(y_test,y_pred)
cm_ma = confusion_matrix(y_test, y_pred)

6 Conclusion
Machine learning has currently become a part of our daily lives, and is a new tech-
nology endowed, to handle security and privacy problems that arise in cloud com-
puting. In recent years, in depth analysis was administered on the protection and
privacy problems of cloud computing. Therefore, security and privacy become terri-
bly essential and vital and are to be addressed. For future work, it’s essential for the
researchers to deeply investigate different science primitive’s solutions for security
attacks in cloud computing. A mixed protocol technique will scale back the compu-
tation overhead on the protection and privacy protective solutions. Moreover, cus-
142 | Sachi Nandan Mohanty, Gouse Baig Mohammad, Sirisha Potluri, P. Ramya, P. Lavanya

tomization of the privacy and security protocols for machine learning is additionally
a noteworthy and open analysis space to develop a viable resolution. Research is
looking forward to perform their analysis within various applications of machine
learning such as space of astronomy, natural philosophy, nucleonics, physical sci-
ence, magnetism, machines, technology, mechanics, electro hydrokinetics, signal
process, power, energy, bioinformatics, economy, and finance.

7 References
[1] Dillon T, Wu C, Chang E, Cloud Computing: issues and challenges, 2010 24th IEEE International
Conference on Advanced Information Networking and Applications, Perth, WA, Australia, 2010,
pp. 27-33, doi: 10.1109/AINA.2010.187
[2] Wang L, Laszewski V, Younge A, Cloud Computing: a perspective study, New Gener. Comput,
2010, 28, 137–146, https://doi.org/10.1007/s00354-008-0081-5
[3] Naveen G, Ayushi G, Survey on machine learning based scheduling in cloud computing, Pro-
ceedings of the 2017 International Conference on Intelligent Systems, Metaheuristics & Swarm
Intelligence, Association for Computing Machinery, New York, NY, USA, 2017, 57–61,
DOI:https://doi.org/10.1145/3059336.3059352
[4] Umer A, Muhammad M, Shah S, Amin R, Shaukat W, Syed R, Suh Y, Piran J, A review of machine
learning algorithms for cloud computing security, Electronics, 9, 2020, no. 9: 1379,
https://doi.org/10.3390/electronics9091379
[5] Subramanian K, Tamilselvan L, A focus on future cloud: machine learning-based cloud securi-
ty, SOCA, 2019, 13, 237–249, https://doi.org/10.1007/s11761-019-00270-0
[6] Bhamare D, Salman T, Samaka M, Erbad A, Jain R, Feasibility of supervised machine learning
for cloud security, 2016 International Conference on Information Science and Security (ICISS),
Pattaya, Thailand, 2016, pp. 1-5, doi: 10.1109/ICISSEC.2016.7885853
[7] Nassif A, Talib M, Nasir Q, Albadani H, Dakalbab F, Machine learning for cloud security: a
systematic review, IEEE Access, 2021, vol. 9, pp. 20717-20735, doi:
10.1109/ACCESS.2021.3054129
[8] Han Q, Meikang Q, Zhihui L, Selective encryption on ECG data in body sensor network based on
supervised machine learning, Information Fusion, 2020, Volume 55, Pages 59-67, ISSN 1566-
2535, https://doi.org/10.1016/j.inffus.2019.07.012
[9] Potluri S, Mangla M, Satpathy S, Mohanty S N, Detection and prevention mechanisms for DDoS
attack in cloud computing environment, 2020 11th International Conference on Computing,
Communication and Networking Technologies (ICCCNT), Kharagpur, India, 2020, pp. 1-6, doi:
10.1109/ICCCNT49239.2020.9225396
[10] Karimipour H, Dehghantanha A, Parizi R, Choo K, Leung H, A deep and scalable unsupervised
machine learning system for cyber-attack detection in large-scale smart grids, IEEE Access,
2019, vol. 7, pp. 80778-80788, doi: 10.1109/ACCESS.2019.2920326
[11] Wu W, Alvarez J, Liu C, et al., Bot detection using unsupervised machine learning, Mi-
crosystTechnol, 2018, 24, 209–217, https://doi.org/10.1007/s00542-016-3237-0
[12] Çavuşoğlu Ü, A new hybrid approach for intrusion detection using machine learning meth-
ods, ApplIntell, 2019, 49, 2735–2761, https://doi.org/10.1007/s10489-018-01408-x
[13] Potluri S, Achyuth S, Elham T, Mohanty S N, IOT enabled cloud based healthcare system using
fog computing: a case study, Journal of Critical Reviews ISSN- 2394-5125, 2020, Vol 7, Issue 6,
PP: 1068-1072, doi: 10.31838/jcr.07.06.186
 Next Generation Cloud Security: State of the Art Machine Learning Model | 143

[14] Mahmud R, Fernando K, Rajkumar B, Cloud-fog interoperability in IoT-enabled healthcare

solutions, ICDCN '18: Proceedings of the 19th International Conference on Distributed Compu-
ting and Networking, 2018, Article No.: 32 Pages 1–10,
https://doi.org/10.1145/3154273.3154347
[15] Subramanian E, Tamilselvan L, A focus on future cloud: machine learning-based cloud security,
Service Oriented Computing and Applications, 2019, 13, 10, 10.1007/s11761-019-00270-0
[16] marwan m, kartit a, ouahmane h, security enhancement in Healthcare Cloud using Machine
Learning, Procedia Computer Science, 2018, Volume 127, Pages 388-397, ISSN 1877-0509,
https://doi.org/10.1016/j.procs.2018.01.136
[17] Sen J, Security and privacy issues in cloud computing, ArXiv, abs/1303.4814, Khanghahi N,
Ravanmehr Reza, Cloud Computing Performance Evaluation: Issues and Challenges, Interna-
tional Journal on Cloud Computing: Services and Architecture (IJCCSA) ,2013, Vol.3, No.5, 3.
10.5121/ijccsa.2013.3503
[18] Sinha N, Khreisat L, Cloud computing security, data, and performance issues, 2014 23rd Wire-
less and Optical Communication Conference (WOCC), 2014, Newark, NJ, USA, 2014, pp. 1-6,doi:
10.1109/WOCC.2014.6839924
[19] Lehrig S, Eikerling H, Becker S, Scalability, elasticity, and efficiency in cloud computing: a
systematic literature review of definitions and metrics, 10.1145/2737182.2737185, Rashidi B,
Sharifi M, Jafari T, A Survey on Interoperability in the Cloud Computing Environments,
I.J.Modern Education and Computer Science, 2013, 6, 17-23, 10.5815/ijmecs.2013.06.03
[20] Potluri S, Subbarao K, Improved quality of service-based cloud service ranking and recommen-
dation model, TELKOMNIKA Telecommunication, Computing, Electronics and Control, 2020,
Vol. 18, No. 3, pp. 1252~1258 ISSN: 1693-6930, DOI: 10.12928/TELKOMNIKA.v18i3.11915
[21] Potluri S, Subbarao K, A hybrid PSO based task selection and recommended system for cloud
data, Test Engineering and Management, 2020, Vol-83, ISSN: 0193-4120 PP: 10210 – 10217
[22] Potluri S, Subbarao K, Hybrid self-adaptive PSO and QoS based machine learning model for
cloud service data, International Journal of Control and Automation, 2020, Vol. 13, No. 2s, pp.
36 - 50, ISSN: 2005- 4297
[23] Nguyen T, Berder O, Sentieys O, Energy-efficient cooperative techniques for infrastructure-to-
vehicle communications, IEEE Transactions on Intelligent Transportation Systems, 2011, vol.
12, no. 3, pp. 659-668, doi: 10.1109/TITS.2011.2118754
[24] Potluri S, Varshith K, Software virtualization using containers in google cloud platform, Inter-
national Journal of Innovative Technology and Exploring Engineering (IJITEE), 2019, ISSN: 2278-
3075, Volume-8 Issue-7
[25] Nguyen T, Berder O, Sentieys O, Energy-efficient cooperative techniques for infrastructure-to-
vehicle communications, IEEE Transactions on Intelligent Transportation Systems, 2011, vol.
12, no. 3, pp. 659-668, doi: 10.1109/TITS.2011.2118754
[26] Challita S, Zalila F, Gourdin C, Merle P, A precise model for google cloud platform, 2018 IEEE
International Conference on Cloud Engineering (IC2E), 2018, Orlando, FL, USA, pp. 177-183, doi:
10.1109/IC2E.2018.00041
[27] Shabani I, Kovaçi A, Dika A, The benefits of using google cloud computing for developing
distributed applications, Journal of Mathematics and System Science, 2015, 5, 156-164, doi:
10.17265/2159-5291/2015.04.004
[28] Butt A, Mehmood M, Shah H, Amin R, Shaukat W, Raza M, Suh Y, Piran M, A review of machine
learning algorithms for cloud computing security, Electronics, 2020, 9, 1379,
https://doi.org/10.3390/electronics9091379
[29] Li K, et al., Assessment of machine learning algorithms in cloud computing frameworks, 2013
IEEE Systems and Information Engineering Design Symposium, Charlottesville, VA, USA, 2013,
pp. 98-103, doi: 10.1109/SIEDS.2013.6549501
144 | Sachi Nandan Mohanty, Gouse Baig Mohammad, Sirisha Potluri, P. Ramya, P. Lavanya

[30] Chiba Z, Abghour N, Moussaid K, Amina E, Mohamed R, Intelligent approach to build a deep
neural network based IDS for cloud environment using combination of machine learning algo-
rithms, Computers & Security, 2019, Volume 86, Pages 291-317, ISSN 0167-4048,
https://doi.org/10.1016/j.cose.2019.06.013
[31] Bhamare D, Salman T, Samaka M, Erbad A, Jain R, Feasibility of supervised machine learning
for cloud security, 2016 International Conference on Information Science and Security (ICISS),
Pattaya, Thailand, 2016, pp. 1-5, doi: 10.1109/ICISSEC.2016.7885853
[32] Gulenko A, Wallschläger M, Schmidt F, Kao O, Liu F, Evaluating machine learning algorithms for
anomaly detection in clouds, 2016 IEEE International Conference on Big Data (Big Data), Wash-
ington, DC, USA, 2016, pp. 2716-2721, doi: 10.1109/BigData.2016.7840917
Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M.
Vaishnavi, K. Sahaja
Secure Intelligent Framework for VANET: Cloud
Based Transportation Model
Abstract: Vehicular cloud computing has a prodigious impact on information tech-
nology solutions. According to VCC the unexploited computing resources are park-
ing vehicles, vehicles stuck in congestion, parked cars in any place. VCC architec-
ture has three layers. Some of the developments that are taking place in the
automobile industry for the last few decades have been very interesting and resolu-
tion driven. A lot of safety applications, convenience and commercial applications
are proposed for the VANET’s by researchers and by the US department of transpor-
tation. The vehicles connect together through a network to form a vehicular cloud
which will offer space for storing, computing resources, sensor readings as an on-
demand service to clients. Traffic is becoming a daily and even growing phenome-
non which spoils the valuable time, fitness and energy of human. There is a need of
efficient computational, supervising and guiding solutions for the citizens to make
them to get rid of traffic related issues and problems. Finding an appropriate park-
ing lot within the area on the brink of university, shopping complex, or a commer-
cial complex in big and medium cities would enjoy the assistance of an automatic
parking management utility.

Keywords: Vehicular Cloud Computing, Intelligent Transportation Systems, Cloud

Security, Cloud Privacy, Vehicular Ad Hoc Network, Internet of Things

||
Sirisha Potluri, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI Foundation
for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana 501203, India,
[email protected]
Gouse Baig Mohammad, Department of CSE, Vardhaman College of Engineering, Hyderabad,
Telangana 501218, India, [email protected]
Sachi Nandan Mohanty, Department of Computer Engineering, College of Engineering Pune,
Wellesley Rd, Shivajinagar, Pune, Maharashtra 411005, India, [email protected]
M. Vaishnavi, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI Foundation
for Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana 501203, India,
[email protected]
K. Sahaja, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI Foundation for
Higher Education, Donthanapally, Shankarapalli Road, Hyderabad, Telangana 501203, India,
[email protected]

https://doi.org/10.1515/9783110732573-008
146 | Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K. Sahaja

1 Introduction to VANET
Some of the developments that are taking place in the automobile industry for the
last few decades have been very interesting. Vehicles today are considerably more
fuel-efficient than they were before. Even in today’s time the vehicles are prone to
accidents due to mist, smoke, dust and various other death-traps on the road but
above all of that they are much more vulnerable to accidents that are taking place
due to human errors.
However much of these fatal human errors are about to reduce because the au-
tomobile industry can been working vigorously for many year to put different kinds
of sensors in the cars and connect them to an on-board computer system. With the
development in the technology telecommunications and broadcastings has now
become potential to connect vehicles to each other through wireless technologies to
enable them to interconnect and work together. Autonomous vehicles are being
supported by the governments in this present age.
A lot of funding is provided for research on driver-less cars from various gov-
ernments especially from UK. This funding is bringing about the merging of various
companies together which not only include the companies from automobile sector
but also from information technology, telecommunications to work together.
Many car manufacturers such as Ford, Hyundai, and Tata etc. have released ap-
plication development for their platforms by making their application program in-
terface- API available to other developers. The developers submit their apps and
approve them before making them available for download. The long term evolution-
LTE connectivity ensues to serve and improve the in-vehicle presentation services
by providing entrance and access to high-speed internet, streaming movies, naviga-
tion control, music applications, and live television and so on. The other feature is
commercial and profitable which offers location-based services and ads to the vehi-
cle travellers. The LTE connectivity helps in bringing the internet not only to the
vehicle but also makes the vehicle a part of the internet paving the way for the ve-
hicular ad hoc networks (VANETs).
The true future prospective of the connected vehicle can only be realized when
vehicles are interconnected to each other. This network that is formed by the inter-
connection of vehicles is referred to as VANET. This shift in vehicular technology
will bring about a new era of innovation and will open a huge range of application
areas that can help in improving road safety and reducing accidents on roads.
VANETs are considered to be significant because of their huge potential and
several applications that are being made accessible. VANETs not only offer strong
safety enhancements but also provide a lot of commercial openings and opportuni-
ties. Wireless access in the vehicular environment (WAVE) delivers the basic radio
standard for dedicated short range communication (DSRC) in VANETs.
 Secure Intelligent Framework for VANET: Cloud Based Transportation Model | 147

Vehicles use DSRC radios to connect and communicate with each other in two
ways namely vehicle-to-vehicle communication (V2V) and vehicle-to-infrastructure
communication (V2I). The communication range of dedicated short range commu-
nication lies in the range between 300 to 1000 m. With the addition of 4G/LTE con-
nectivity in vehicles it has become obvious that connected and networked vehicles
have the facility to share data by means of the cloud. The gains of using cloud in
connected vehicles are enormous as the opportunities for using applications are
virtually limitless.
DSRC identifies one main type of message which is called as the basic safety
message (BSM) that a vehicle communicates and transmits every 100 milli seconds.
In basic safety message the vehicles broadcast the details of their location, speed,
acceleration, and several other useful parameters which can help other vehicles in
avoiding collisions and generate warnings for having improved safety measures.
These messages also help the vehicles and auto mobiles in maintaining the list of
their neighbours.
The navigation systems has been made available but the exact real-time loca-
tion information of other vehicles is not yet available. This is an area in the future
that can enable the driver to drive safely on roads even if there is nil visibility or
when the driver is not capable to drive the vehicle. Therefore, the vision is to gener-
ate a mutual understanding among all vehicles on the road that can lead to driver-
less cars.
All the vehicles can be programmed and automated by using software to trans-
mit and share their location and other related important information whenever they
observe an unexpected change in speed or position while driving. However, the
bandwidth in DSRC is quite limited, so when the vehicles try to transmit their loca-
tion or position and speed information there is a possibility of having a jammed
channel. The DSRC and 4G/LTE connectivity in vehicles can support each other and
answer the bandwidth problem to give great quality, consistent, reliable, and secure
information to VANET users [1-5].

2 Constraints of Cloud
Cloud computing offers on demand services to the customers using efficient re-
source usage and allocation based policies and protocols by matching with the ser-
vice level agreement. Service level agreement commonly includes various segments
like measuring the performance, managing the problems, customer satisfaction,
recovery from failures and finally termination of the agreement. SLA’s are consist-
ently checked and validated using which it creates a better communication channel
between consumer and the provider. SLA establishes the Quality of Service (QoS)
agreement between service-based system providers and users. With a violation of
148 | Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K. Sahaja

SLA, the provider must pay penalties. Continuous performance check is required in
order to ensure that these CSPs are providing services as mentioned in SLAs. SLA is
consistently met, these agreements are frequently designed with specific lines of
demarcation and the parties involved are required to meet regularly to create an
open forum for communication. Cloud computing is using collection of resources
and all these are supported by the interface and infrastructure concept of distribut-
ed and cloud computing environment. Service level agreement can be viewed as a
document between customer and provider and is mainly based on services and not
on customers [6-10].

3 Applications of VANET
A lot of safety applications, convenience, business and commercial applications are
proposed for the VANET’s by researchers and by the US department of transporta-
tion. Some of these applications are discussed briefly below. With numerous vehi-
cles on the roads having such as lot of computing power and sensors, it’s apparent
and logical to utilize all this data to make a highly mobile unplanned networks [11].
Some of the applications that are recommended for VANETs are as follows.

3.1 Safety applications

Some of the safety applications include:
– Sending notifications if any crashes occur.
– Threats that might happen on the roads due to bad, slippery or wet road condi-
tions in cases of heavy rains.
– Traffic breaking warnings and curve speed warnings in case of fast speed or
negligence.
– Emergency electronics traffic control or any pre-crash feel and co-operative
forward collision warnings.

Some of the safety precautions might also include warning messages that are given
to drivers in order to inform them about approach of any emergency vehicles.

3.2 Convenience applications

Some of the convenience applications include:
– Navigation and personal routing is provided to reach destination, stoppage
advice is given in case of heavy traffic.
 Secure Intelligent Framework for VANET: Cloud Based Transportation Model | 149

– Toll group and parking availability information is provided for convenience of

drivers.
– Road and weather are often monitored by sharing the information from on-
board vehicle sensors.
– The vehicular network can turn into the crisis and emergency transmission
device.

The connected and networked vehicles can play a really important role in situations
such as power disruption and network breakdown as they need on-board batteries
and lots of sensors including cameras which provide valuable images and SOS calls.

3.3 Commercial applications

Some of the commercial applications include:
– Location-based facilities and services like advertisements, announcements and
entertainment which include data/information/video relay, social networking
informs.
– By using vehicle to vehicle communication and vehicle to infrastructure com-
munication information such as vehicle positions and information on road con-
ditions can be easily communicated.

Furthermore, the driving force may have to form life-saving decisions that are sup-
ported by the knowledge received such as hand brake light, onward collision warn-
ing. It is therefore extremely important that the privacy, security, reliability, and
integrity of the messages are ensured and safeguarded in order that actions are
taken immediately after the knowledge is received.

4 VANET- Cloud Based Approach

The great amount of unused and idle space for storing and computing power includ-
ing on-board sensors provide a singular opportunity to use vehicles as a wireless
sensor network. The vehicles are going to be equipped and furnished with not only
sensors to live temperature, rainfall, humidity, wind, etc., but also with cameras.
This alongside high speed (4G/LTE) connectivity are often an excellent benefit for a
spread of cloud based vehicle applications [12].
150 | Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K. Sahaja

4.1 VANETs within the cloud

The connected and networked vehicles can link with the cloud and can upload
their data which include sensor readings and traffic information that can be
downloaded by other vehicles that are following behind schedule. This enables
the users to share data unknown and preserve security and privacy of information
that is being provided as shown in fig. 1. The application servers that are embed-
ded within the cloud can collect and add the knowledge coming in from vehicles
and then bring together that information to form announcements and declarations
for traffic, road conditions and surroundings, accidents and weather supported
vehicle sensor data. This progresses the reliability of the knowledge by filtering
out the biased or dishonest information gauged by the amount of users or con-
sumers reporting such information.

Traffic
information

Broadcasting Weather
utilities information

VANETS in the cloud

Infotainment
Toll collection
services

Fig. 1: VANETs in the cloud

 Secure Intelligent Framework for VANET: Cloud Based Transportation Model | 151

4.2 Vehicular cloud

The vehicles join and connect together through a network to form a vehicular cloud
which will offer space for storing, computing resources, sensor readings and inter-
pretations as an on-demand service to clients. The clients who demand such ser-
vices may be individuals, organizations, corporations and government departments
like the meteorology, climatology department, police department or highway au-
thorities as shown in fig. 2. Moreover, with cameras on board vehicles and high-
speed connectivity, the police or disaster services can even appeal images from
vehicles for a neighbourhood that they need to watch like the location of an acci-
dent [13].

Data storage

Vehicular data Computing

analytics resources

Vehicular cloud

Multimedia Weather data

pooling collection

Fig. 2: Vehicular cloud

152 | Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K. Sahaja

4.3 Private vehicular cloud

The computing resources, storage facilities of the vehicle are often used by the own-
er as a personal cloud in which the user attaches to the web and uploads all of the
data into a private cloud as shown in fig. 3. By storing all of the data into a private
cloud one can get ownership of data and view it whenever it is required [14].

Data storage

Secured
confidential Computing
information and resources
systems

Private vehicular cloud

Privately shared
virtualized Hotspot
resources

Fig. 3: Private vehicular cloud

5 Autonomous/Self-Governing Vehicle
The autonomous or self-governing vehicle may be a certainty today through Google
car and it’s only logical to mention that the technology and expertise will improve
further and decrease in cost to become a feature within the near future. However,
 Secure Intelligent Framework for VANET: Cloud Based Transportation Model | 153

it’ll take a while for it to develop a driver-less feature on a massive scale. Also, it’s
possible that very similar to control, this might become a feature that would be sim-
ulated and activated on the motorway or highway. Autonomous vehicles would
wish almost no input from the driving force. This is able to be possible with tons of
hardware like sensors, cameras, radars, locators and laser illuminated detection and
ranging (LIDAR) to sense approaching obstacles and hurdles. Image processing
practices are applied to pictures from the cameras to acknowledge, classify and
identify the various shapes and features. This permits the vehicle to sense the up-
coming hurdle or obstacle and spontaneously apply breaks or reduce speed far more
rapidly than the driving force would ever be ready to. The utilization of those and
other technologies therefore safeguards that human errors are often abridged to an
outsized extent and, as an outcome, roads are regularly made much safer. This is
able to enable the driving force to be more dynamic by doing something like read-
ing, evaluating and browsing on the web or making a call.
Autonomous vehicles aren’t only an opportunity but an essential for safe motor-
ing. Subsequent step in autonomous or self-governing vehicles is vehicle swarm [15].

6 Vehicle Platoon/Swarm
Travelling at great speeds take along more problems however, these problems are
often abridged if the vehicles form a gaggle then travel during the creation on high-
ways and motorways. This may have two benefits:
– Vehicles are synchronized and corresponded (i.e., vehicles travel during a line
and maintain a hard and fast distance from one another and therefore the vehi-
cle ahead informs the cars behind if they ought to reduce speed or change lanes
beforehand , then all of them can roll in the hay in tandem)
– The randomness within the immediate neighbourhood of the vehicles is re-
duced, which makes things much simpler and harmless.

Safe road trains for the environment (SARTRE) is an EU-endow more than 6 million
dollar project that has exhibit this idea together with Volvo, Ricardo and among
others. The purpose of the SARTRE development was to develop a road train or pla-
toon conception in vehicle traffic where a lead vehicle driven by well-informed driv-
er takes control of a platoon of vehicles that enter a semi-independent state. The
SARTRE project established successfully that vehicle platoons aren’t only safe but
are often very efficient in reducing congestion and improving safety. The vehicle
swarm are often imaginary by thinking of the motorway/highway as a conveyer belt,
but where a conveyer belt moves with a continuing steady speed, a gaggle of vehi-
cles on the auto route can move with any set sustained speed. This will be probable
by agreeing like-minded vehicles, i.e., vehicles that shall travel at roughly an equiv-
154 | Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K. Sahaja

alent speed, to team and form groups or a “vehicle swarm”. Ricardo, a British engi-
neering firm convoluted within the SARTRE project, considers that vehicle platoon-
ing is feasible with fully autonomous or self-governing vehicles [16].

7 Applications of Vehicular Cloud Computing

7.1 Parking lot data or information based cloud
Topical US-DOT indicators reveal that the registered vehicular convoy on American
streets and roadways is nearly 256 million vehicles are strong and rising gradually.
Also, statistics disclose that most of those vehicles spend numerous hours per day
parked for the duration of a parking garage, parking region, or driveway. Allow us
to consider a small company paying a couple of hundred people and given that IT
services. We leave carpooling and vehicles remain parked in the parking lot. In
those vehicles, the algorithmic resources have no use and persist inactive for hours.
The corporate can give an intention for the employees who will rent the resources
for the VC formation. Hence, the contributing and take part vehicles within the park-
ing lot create a computer vehicle cluster and supply ample services for secured data
storage [17].

7.2 Dynamic and self-motivated traffic light management

Todays, thanks to amassed variety of vehicles on the roads, the traffic is flattering a
daily phenomenon which wastes the valuable time, resources, energy and fitness of
the human. One among the finest solutions to worn out this issue is to allocate the
appropriate quantity of resources instead of pre-allocating of massive resources as a
basis for the worst condition. Allow us to think through an occasion like a football
match that’s appeared by thousands of individuals, where a traffic congestion may
occur at the top of the diversion. Even though range of studies and trainings are
accomplished to deal with this subject by leveraging VANETs and IT’S, they are
unable to testimony traffic problems rapidly and typically cannot provide a traffic
alleviation plan. VCC is in a position to present an additional well-organized and
cost-effective thanks to solve the traffic jam by providing the definite resources from
the available vehicles contributing within the traffic and connecting them in finding
a solution unconventionally without expecting officials react [18].
 Secure Intelligent Framework for VANET: Cloud Based Transportation Model | 155

7.3 Optimizing and enhancing traffic signals

Traffic signals set the sign cycle length and consequently the green phase lengths.
Signal system optimization is presently take place off line at either out of the way
connection or the passageway level. The time periods are well-defined by the timing
strategies surely time periods, such as the workaday morning or afternoon greatest
peak hours. One among the drawbacks of this method is that it requires data on
traffic whirling movements that are frequently collected to make guaranteed that
the signal timing strategies are appropriate for the present traffic capacity condi-
tions. Another difficulty is that this plan does not cope well with indefinite changes
in traffic circumstances and environments. Thus, scan exploits the signal system
enactment by making active use of a vehicular network.

7.4 Road safety communication and message

New cars have embed detecting devices for skilful and safe operation. The cameras
help the driving force by chasing the lines on the road and support them in staying
within the path. Thus, cars have a sensor node, and a VC can form dynamic with a
big wireless sensor network. Vehicles review the sensors of other cars in close near-
ness to extend the loyalty and obtain a valuation of the probable road risk ahead,
the road conditions, speed breakers, dumps, and snow. However, this VANET de-
sign will not give us the direction of safety measures and a quicker solution.

7.5 VC in developing countries

Due to lack of skilled centralized decision support systems and provision, the idea
of VC are going to be very significant in emerging countries also. Moreover, VCs will
play a significant role in making a vast number of computing resources reachable
through a vehicular network by using many computing applications dynamically,
which aren’t likely to use with the existing infrastructure.

7.6 Managing parking facilities

Finding an appropriate parking lot within the area on the brink of university, shop-
ping complex or in big cities would enjoy the assistance of an automatic parking
controlling utility. Recently, the recognised solutions reinforced a unified style from
single parking garages and parking meters have been joined then spread to the
community. Today, the vehicle drivers are capable to find a parking lot by means of
their smartphone applications (e.g., Smart park, Park Me and Park Mate). However,
these applications have some disadvantages that haven’t been addressed, such as:
156 | Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K. Sahaja

covering the incomplete point of the planet and therefore the need for accessing
internet. The vehicles that form the VC during a specific area will accomplish real
time data of accessible parking spaces and initiate drivers to the most appropriate
position [19].

8 Characteristics of VANET
The characteristics of VANETs consist of high mobility and regular change in topol-
ogy. Some of the characteristics of VANETs are discussed below:

8.1 High mobility

VANETs have great mobility. Having high mobility in VANETs essentially helps in
playing a very significant role in the demonstrating of VANET protocol. Every node
that exist in a VANET moves on very great speed which allows the great mobility of
nodes and decreases the communication time in the vehicular network.

8.2 Driver safety

VANETs can support in educating the driver safety, improve the passenger ease
level and progress the traffic flow. One of the foremost gain of VANETs is that the
vehicles can connect and communicate directly with each other. It allows a number
of submissions to connect and communicate among different nodes such as the
Road side units (RSU) and On-board units (OBU).

8.3 Dynamic network topology

The topology of VANETs can change quickly based upon the vehicle speed and great
mobility. The changes that are taking place in the flexibility of the vehicles are creat-
ing the VANETs more horizontal to various kinds of attacks and it is also becoming a
very hard job to distinct the suspected vehicles.

8.4 Frequent network disconnection

The regular suspension and disconnection of the VANETs are become high-speed
measure among vehicles and various other usual conditions such as weather. A
 Secure Intelligent Framework for VANET: Cloud Based Transportation Model | 157

large amount of vehicles on the road can also lead to the regular disconnection in
the service.

8.5 No power constraints

There is certainly no power constraint in VANETs .This allows the vehicle to deliver
constant power to the on-board units by the means of a long-lasting battery.

8.6 Network strength

The network strength of VANETs be influenced by the flow of the traffic on the
roads. If there is a case of traffic congestion the network strength can be very high
and if there is completely no traffic it can be low. The network can be superior near-
by highways and close to the entry and exit locations of the city.

8.7 Large computational processing

The nodes that are existing in the VANETs are vehicles which can be inserted with
the sensors and some of the other computational devices such as processors, anten-
na and GPS. All of these resources will need to have a great quantity of computa-
tional capacity of the node in the VANET and provide healthier wireless communi-
cation to obtain specific information relating to position, speed and direction.

9 Security Services of VANET

The security and safety of the VANETs is an important issue for providing safety to
the person along for the ride and drivers. The security services are as follows:
– Availability of the service
– Confidentiality of the service
– Authentication of the service
– Data integrity of the service
– Nonrepudiation of the service

9.1 Availability of the service

The feature of Availability plays an extremely crucial role in the security of the
VANET because it will ensure that the network will always remain functional in a lot
158 | Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K. Sahaja

of faulty and unpredicted conditions. This feature of availability has become ex-
tremely prone to some of the dangerous attacks as compared to some of the other
security services that are available. Hence a lot of trust-based and cryptography
practices can be used to secure the VANETs from these outbreaks.

9.2 Confidentiality of the service

Confidentiality guarantees that the data and records can be accessed only by the
dedicated user while it makes impossible for unauthorized users to access the pri-
vate and confidential information that communicates to the designated user.

9.3 Authentication of the service

The authentication shows an important role to guarantee the safety of the VANET. It
prevents the vehicles from malicious activities that might take place in network. In
order to avoid the network from dangerous outbreaks it is necessary that the au-
thentication of users and messages which permit through the network is required.

9.4 Data Integrity of the service

The feature of Data integrity helps to ensure that the content of the message is not
changed or altered in any sense during the process of transmission.

9.5 Nonrepudiation of the service

The feature of nonrepudiation is very important as it ensures the validity and makes
sure that the sending and receiving entities cannot reject its transmission and recep-
tion in case of any disagreement.

10 Threats and Attacks of VANETs

The VANETs go through different types of threats and attacks and the destruction
caused by these attacks can cause distress to various other applications. The ad-
verse impact of some of the attacks effects many processes and measures. Some of
these applications include safety, security, comfort, and infotainment.
 Secure Intelligent Framework for VANET: Cloud Based Transportation Model | 159

10.1 Attack on communication

In VANETs it is difficult to achieve secure communication because the VANETs se-
curity goes through many threats such as certificate duplication attack, eavesdrop
attack, and location confidentiality or privacy attack.

10.1.1 Certificate replication attack

In this type of attack attacker may use false certificates and keys of the other users
as an evidence of authentication without getting tracked. The main purpose of such
a type of attack is to create confusion and make it harder to identify the malicious
vehicle.

10.1.2 Eavesdrop attack

This type of attack finds the confidential and personal information when a non-
registered vehicle uses a valid and legal certificate to collect all of the valuable in-
formation of the vehicles such as user ID, vehicle location etc.

10.1.3 Privacy attack

This type of attack includes different kinds of outbreaks on privacy-preserving

schemes that can include something like tracking of a vehicle. For suppose the at-
tacker might try to use the location of the target vehicle, ID, significant and creden-
tial to initiate alternative attack without getting tracked.

10.2 Attack on safety applications

These attacks are associated and related to the channel allocation

10.2.1 Denial of service attacks

This is one of the most common outbreaks in VANETs, the attack occurs when the
fraudulent vehicle directs multiple messages which blocks all potential ways of
communication. The outbreak can be accomplished by many attackers simultane-
ously.
160 | Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K. Sahaja

10.2.2 Jamming attack

Jamming attack is one of the unsafe attacks in VANETs security applications and it
takes place when one of the suspected vehicle tries to interrupt the streaming com-
munication through various kinds of techniques such as taking in heavy power with
equal frequency choice and alert injection.

10.2.3 Platooning attack

It occurs when a large number of unreliable vehicles are located in the same zone or
they try to move forward together with the purpose of causing some kind of trouble
and removing reliable nodes from the network operation that are restricted for using
similar kind of bandwidth.

10.2.4 Betrayal attack

This kind of attack takes place when a reliable vehicle enters into a malicious node
and starts in sending fake messages by the node.

10.3 Attack on infotainment applications

The attacks on infotainment applications tend to talk about the issues that are relat-
ed to the comfort level and safety measures of the passengers.

10.3.1 Illusion attack

Data is received from the antenna and some of the information is collected from the
sensor and passed onto the hardware components.it helps in creation of warning
messages of the road conditions which help in creation of an illusion of the vehicles
that are nearby.

10.3.2 Inject message attack

This type of attack takes place when an unreliable vehicles tries to create a fake
message or duplicate the same message or try to create a new type of message by
passing in some malicious information and modifying the original message in the
network and acts as master node in the inter vehicle communication network.
 Secure Intelligent Framework for VANET: Cloud Based Transportation Model | 161

11 Security and Privacy Related Issues in VANET

Various security schemes have been used to safeguard privacy based on the security
mechanisms they use, these schemes can be classified as:
– Pseudonyms coupled with public key infrastructure (PKI)
– Trust-based schemes
– Group signatures

11.1 Pseudonyms coupled with PKI-based schemes

Pseudonyms preserve and protect the privacy of the users. It hides the original iden-
tity of the user and avoids any kind of link ability of the identity with a pseudonym.
However pseudonyms can reveal the true identity in case the user commits any type
of illegal activity hence providing traceability in case of any wrong actions.
The disadvantage of the pseudonym scheme is that the On Board unit has thou-
sands of pseudonyms stored which need to be refreshed on a timely basis. The
pseudonyms are continuously changed in order to maintain the privacy of the user.
The pseudonyms need to maintain CRL. CRL stands for Certificate Revocation
List. This list keeps information of all the vehicles that are unreliable and untrust-
worthy. This list needs to be checked continuously which can be both time consum-
ing as well as resource consuming. The certificate authority termed CA links the
pseudonyms to the vehicles and hence it must be kept secure in order to maintain
the privacy of the users.

11.2 Trust based schemes

Trust-based schemes have been proposed for MANETs. The same has been applied
to VANETs as well. In VANETs it includes a large number of nodes and as the vehi-
cles move very quickly the topology keeps on changing quickly. So it becomes diffi-
cult to maintain the trustworthiness of the vehicles.
In VANETs the purpose is to increase road safety and hence the decision-
making process needs to be very fast due to the high speeds and limited amount of
time involved. Trust establishment is done through central authority or by means of
security infrastructure. Security infrastructure requires development of a trust score
dynamically. The trust score needs to be maintained and checked.
Protection against inside attackers gets difficult in trust based schemes when a
group of vehicles combine together to form a network. In order to solve this issue
data centric trust schemes have been proposed. In this a false node can be deter-
mined easily based upon the information exchange of whether they are forwarding
the wrong kind of data. This kind of false node can be reported and ignored.
162 | Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K. Sahaja

11.3 Group signatures

For a group of vehicles travelling together at the same speed and direction the Group
signatures for the VANETs have been proposed .This proposed model helps in issuing
messages and sign the messages with the group signatures in behalf of the entire
group. By forming groups privacy can be achieved. The vehicle to vehicle transmission
is reduced by means of periodic broadcasts by a single member of the group.

12 Security and Privacy in VCC

The vehicular internet will motivate the longer term of vehicular technology and
intelligent transportation systems (ITS). Whether its road safety, infotainment,
broadcasting device or self-driving cars, the vehicular internet will lay the motiva-
tion for the longer term of road travel. Governments, organisation bodies and corpo-
rations are pursuing driver-less or self-driving cars as they’re considered to be more
trustworthy than humans and, therefore, harmless and safer. The vehicles today
aren’t just a way of transportation but also are well-appointed with a decent range
of sensors that provide valuable data for decision making. If vehicles are allowed to
share data that they collect with other vehicles or facilitate for decision-making and
harmless driving, and by this means form a vehicular network. However, there are
tons at stake in vehicular networks if they’re negotiated and compromised. With the
risks so high, it’s vital that the vehicular networks are protected and made strong to
any attack or attempt which will have severe result. The vehicular internet also can
be the target of a cyber-attack, which may be overwhelming.
The security and privacy are one among the main characteristics of the vehicu-
lar network, which inaugurates, preserves, and enhances the user’s expectation
within the VC. In particular, the privacy is employed to make sure the communica-
tion and knowledge interchange within the VC which is in reliable environment,
while the safety is active to guard the vehicular network from warning and attacks,
and also from malicious nodes within the network. Olariu et al. considered VCC as a
group of vehicles, during which a group of vehicles in VCC could share the vehicle
resources like computer resources, the web, computing, and communication to form
a predictable cloud computing. Therefore, it leads to cause the same security issues
as CC 7almost the VCC and VANETs are partaking the same security issues.
The core security challenges of VCC includes the following:
– Make sure the secure location because the maximum of the applications depend
upon location information
– Node’s verification and message probity
– Preserve data on cloud from spiteful users
– Privacy of message with cryptography practice.
 Secure Intelligent Framework for VANET: Cloud Based Transportation Model | 163

The security and privacy concerns within the VCC need care from the researchers to
develop a strong algorithm which might be ready to stop VCC from different sorts of
security threats.

12.1 Security services of vehicular cloud computing

There are mainly five main security requirements. They are:

12.1.1 Confidentiality

It certifies that the main data should not be reveal to the outside nodes.

12.1.2 Integrity

It certifies that message should be valid and the contents of the message are not
adjusted during communication process.

12.1.3 Availability

Whenever the vehicle requires some data resources they must be available.

12.1.4 Authentication

In this the VC will be protected by authentication from suspected nodes by checking

the user identity and sender address.

12.1.5 Privacy

The vehicles or passengers will be protected from the attackers by the sensitive and
confidential information.

12.2 Threats of vehicular cloud computing

The demand of using VCC is increasing very year by the challenges faced by it. Here
are some of the threats of Vehicular Cloud Computing.
164 | Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K. Sahaja

12.2.1 Denial of service (DOS)

It is a usual attack on vehicular cloud, these attacks happens to make the services
unavailable to other cloud users.

12.2.2 Identity spoofing

In this attack the registered user identity and security credentials will be misused by
evil user.

12.2.3 Message interferes

This attack happens when the attacker alters the previous messages data to be
transmitted. For instance, if the route is crowded, the attacker modifies the data to
empty the road which can affect the users to change their driving paths.

12.2.4 Information disclosure

In the absence of data privacy this attack takes place by acquiring the useful data of
the system.

12.2.5 Sybil attack

The other vehicular behaviour will get manipulated by these attackers and the vehi-
cle thinks that the information is transmitting from other vehicles. Thus they feel
that there is a crowding on the road, so they impose vehicles to change their way
and leave the road clear.
Vehicular cloud computing security and privacy related concepts are represent-
ed in fig. 4.

12.3 Privacy challenges of vehicular cloud computing

The statistics that’s collected from the vehicle’s sensors has got to be protected and will
only be used after anonymizing it. The worry is that your vehicle should advise you if
you’re over the directive but shouldn’t turn against you, i.e., reporting to the establish-
ments and authorities that you basically went over the speed limit. However, there are
many examples when sharing location information with some cloud service providers
 Secure Intelligent Framework for VANET: Cloud Based Transportation Model | 165

might essentially be beneficial for the user, e.g., insurance enterprises might offer to
measure back the insurance premiums if such details are shared. There are often simi-
lar reasons from other service providers to use vehicle sensor data communally for a
few rewards which will actually encourage contribution from the consumers.

Security
issues
Sybil attack Integrity

Information Confidentiality
disclosure

Vehicular cloud security and

Tampering Availability
privacy

Identify
Privacy
spoofing

Denial of Authentication
service
Threats

Fig. 4: Security and privacy requirements of vehicular cloud computing

13 Solutions to Security Issues

We recommend a digital identity structure for VANETs by seeing that it’s the driving
force who is to be held accountable for the vehicle and not the other way around.
Therefore, we use the ID of the vehicle and therefore the driver to make a standby
digital identity. The two identities are merged into one after verification from two
166 | Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K. Sahaja

authorities, i.e., the vehicle registration authority and therefore the driver license
authority. Each of the two authorities concerns the user token and secret keys which
are then working by the user to get their pseudonym. Therefore, even though the
passkey is compromised, the tokens can only be traced by the two authorities and
no individual authority can disclose truth uniqueness of the user. The performance
of the scheme is enhanced through conditional privacy preservation (ECPP) and
pseudonymous authentication-based conditional privacy (PACP).
Python code for anomaly detection- road condition is given in the following
code listing.

Listing 1: Python code for anomaly detection- road condition to avoid accidents

# Initialize the following variables

P_T_C -->PATH_TO_CKPT,
P_T_L -->PATH_TO_LABEL
N_C -->NUM_CLASSES
#Load the Tensor flow model code into memory
d_graph = tf.Graph()
with d_graph.as_default():
#code to load the model
#Loading label map
l_map = label_map_util.load_labelmap (P_T_L)
catg=l_map_util.convert_label_map_to_categories (l_map, max_num_clas
ses=N_C, use_display_name=True)
c_i = label_map_util.create_category_index (catg)
def load_image_into_numpy_array (image):
#define image and loading details
#To detect images
P_T_T_I_D = '/user/ubuntu/data_set/road_dataset-TF/road_dataset/'
D_T = # set with different road type values
R_L = # set with location values
v_l = []
for loc in R_L:
img_f = open (# open images in read mode)
for entry in img_f:
entry= entry.rstrip ('\n').split ('/')[-1]
v_l.append (entry)
img_f.close()
print(str(len (v_l))
with detection_graph.as_default():
with tf.Session (graph=detection_graph) as sess:
i_t = detection_graph.get_tensor_by_name ('image_tensor:0')
 Secure Intelligent Framework for VANET: Cloud Based Transportation Model | 167

d_b = detection_graph.get_tensor_by_name ('detection_boxes:0')

d_s = detection_graph.get_tensor_by_name ('detection_scores:0')
d_c = detection_graph.get_tensor_by_name ('detection_classes:0')
n_d = detection_graph.get_tensor_by_name ('num_detections:0')
for img_p in T_I_P:
img = Image.open (img_path)
imgs to have shape: [1, None, None, 4]
img_np_expanded = np.expand_dims (img_np, axis=0)
(boxes, scores, classes, num) = sess.run (#detection of boxes,
scores and classes of the images)
vis_util.visualize_boxes_and_labels_on_image_array (
#set the image characteristics)
#Plot the figure using figure and imshow

14 Conclusion
According to our report We conclude that the VANETs becomes very fashionable with-
in the traffic management system, which aims to make sure the security of human
lives on the road and supply comfort to travellers by broadcasting safety messages
among vehicles. As these safety messages are transmit in an open-approach environ-
ment that creates VANETs more permitting to the attacks, a strong security algorithm
must be designed for set about security threats and attacks which could make sure the
secure transmission within the VANETs and VCC.
In this report, we first present the introduction of the VANETs and its applications.
Then, about the VANET and clouds in that we described briefly about VANET within
cloud, vehicular cloud, Private cloud and autonomous vehicle. Second, we’ve dis-
cussed about the vehicle platoon/swarm and then the applications of vehicular cloud
computing. Third, we’ve discussed the VANET characteristics explained followed by
the safety and privacy problems with the VCC intimately. Finally, we’ve presented
some issues associated with VANETs and VCC that are considered as open research
challenges during a vehicular network.
The research management for VANETs basis on the safety and privacy issues like
trust model and cryptography-based technique to validate safety messages. Based on
these techniques, researchers could design a strong security system, which may be
ready to prevent VANETs from different sorts of security threats and attacks. In addi-
tion to VANETs, the VCC residue within the starting stage and expected to supply an
ideal solution to guard the network from different type of threats and improve the
organization of the traffic management system. In our opinion, supported the prevail-
ing VCC procedures, architectures, and practices, an improved algorithm might be
designed to scale back the trust and privacy issues in VCC.
168 | Sirisha Potluri, Gouse B. Mohammad, Sachi Nandan Mohanty, M. Vaishnavi, K. Sahaja

15 References
[1] Coady Y, Hohlfeld O, Kempf J, McGeer R, Schmid S, Distributed cloud computing: applications,
Status Quo, and Challenges, SIGCOMM Comput. Commun. Rev, 2015, 45, 2, 38–43,
DOI:https://doi.org/10.1145/2766330.2766337
[2] Kliazovich D, Arzo S, Granelli F, Bouvry P, Samee U, e-STAB: Energy-efficient scheduling for
cloud computing applications with traffic load balancing, 2013 IEEE International Conference
on Green Computing and Communications and IEEE Internet of Things and IEEE Cyber, Physical
and Social Computing, Beijing, China, 2013, pp. 7-13, doi: 10.1109/GreenCom-iThings-
CPSCom.2013.28.
[3] Nayak J, Naik B, Jena A, Barik R, Das H, Nature inspired optimizations in cloud computing:
applications and challenges, Mishra B, Das H, Dehuri S, Jagadev A, Cloud Computing for Opti-
mization: Foundations, Applications, and Challenges, Studies in Big Data, 2018, vol 39,
Springer, Cham, https://doi.org/10.1007/978-3-319-73676-1_1
[4] Bahga A, Madisetti V, Synthetic workload generation for cloud computing applications, Journal
of Software Engineering and Applications, 2011, Vol. 4 No. 7, pp. 396-410,
doi: 10.4236/jsea.2011.47046
[5] Heap-Yih C, John W, Xiangyu W, An explanatory case study on cloud computing applications in
the built environment, Automation in Construction, 204, Volume 44, Pages 152-162, ISSN
0926- 805, https://doi.org/10.1016/j.autcon.2014.04.010
[6] Potluri S, Subbarao K, Quality of Service based task scheduling algorithms in cloud computing,
International Journal of Electrical and Computer Engineering, 2017, Vol. 7, No. 2, PP: 1088-1095
[7] Potluri S, Subbarao K, Improved quality of service-based cloud service ranking and recom-
mendation model, TELKOMNIKA Telecommunication, Computing, Electronics and Control,
2020, Vol. 18, No. 3, pp. 1252~1258 ISSN: 1693-6930, accredited First Grade by Kemen-
ristekdikti, Decree No: 21/E/KPT/2018 DOI: 10.12928/TELKOMNIKA.v18i3.11915
[8] Potluri S, Subbarao K, A hybrid PSO based task selection and recommended system for cloud
data, test engineering and management, 2020, Vol-83, ISSN: 0193-4120 PP: 10210 – 10217
[9] Potluri S, Subbarao K, hybrid self-adaptive PSO and QoS based machine learning model for
cloud service data, International Journal of Control and Automation, 2020, Vol. 13, No. 2s, pp.
36 - 50, ISSN: 2005- 4297
[10] Potluri S, Subbarao K, Optimization model for QoS based Task scheduling in cloud computing
environment, International Journal of Electrical, Electronics and Computer Systems, 2020, Vol.
18 , No 2, pp. 1081-1088, ISSN: 2502-4752, DOI: 10.11591/ijeecs.v18.i2.pp1081-1088
[11] Whaiduzzaman M, Sookhak M, Abdullah G, Buyya R, A survey on vehicular cloud computing,
Journal of Network and Computer Applications, 2014, Volume 40, Pages 325-344, ISSN 1084-
8045, https://doi.org/10.1016/j.jnca.2013.08.004
[12] Qin Y, Huang D, Zhang X, VehiCloud: Cloud computing facilitating routing in vehicular net-
works, 2012 IEEE 11th International Conference on Trust, Security and Privacy in Computing and
Communications, Liverpool, UK, 2012, pp. 1438-1445, doi: 10.1109/TrustCom.2012.16
[13] Jiau M, Huang S, Hwang J, Vasilakos A, Multimedia services in cloud-based vehicular networks,
IEEE Intelligent Transportation Systems Magazine, 2015, vol. 7, no. 3, pp. 62-79, doi:
10.1109/MITS.2015.2417974
[14] Bitam S, Mellouk A, Zeadally S, VANET-cloud: a generic cloud computing model for vehicular
Ad Hoc networks, IEEE Wireless Communications, 2015, vol. 22, no. 1, pp. 96-102, doi:
10.1109/MWC.2015.7054724
 Secure Intelligent Framework for VANET: Cloud Based Transportation Model | 169

[15] Zhao J, Li Q, Gong Y, Zhang K, Computation offloading and resource allocation for cloud assist-
ed mobile edge computing in vehicular networks, IEEE Transactions on Vehicular Technology,
2019, vol. 68, no. 8, pp. 7944-7956, Aug. 2019, doi: 10.1109/TVT.2019.2917890
[16] Euisin L, Eun-Kyu L, Mario G, Soon Y, Vehicular cloud networking: architecture and design
principles, IEEE Communications Magazine, 2014, vol. 52, no. 2, pp. 148-155, doi:
10.1109/MCOM.2014.6736756
[17] Alazawi Z, Altowaijri S, Mehmood R, Abdljabar M, Intelligent disaster management system
based on cloud-enabled vehicular networks, 2011 11th International Conference on ITS Tele-
communications, St. Petersburg, Russia, 2011, pp. 361-368, doi: 10.1109/ITST.2011.6060083
[18] Yu R, Zhang Y, Gjessing S, Xia W, Yang K, Toward cloud-based vehicular networks with efficient
resource management, IEEE Network, 2013, vol. 27, no. 5, pp. 48-55, doi:
10.1109/MNET.2013.6616115
[19] Hassan A, Luong T, Jin W, Sungyoung L, Saad Q, V-Cloud: vehicular cyber-physical systems and
cloud computing, In Proceedings of the 4th International Symposium on Applied Sciences in
Biomedical and Communication Technologies, Association for Computing Machinery, New
York, NY, USA,2011, Article 165, 1–5, DOI:https://doi.org/10.1145/2093698.2093863
Sirisha Potluri, Sachi Nandan Mohanty, A. D. Sriram Kumar, D.
Maheswari, B. Rahini
Cloud Manufacturing Service: A Secure and
Protected Communication System
Abstract: Cloud computing is the one of the most important technology in the cur-
rent world and it has the high ability to reduce the cost. Cloud computing allows
cloud users to put their information, data and requests on the cloud service provides
like Azure, Amazon AWS and Google suit. Cloud provides computing software like a
development platform where you can create or develop your own applications. The
cloud consumers can use all these resources provided by cloud service provider
from anywhere in the world through the internet facility. Using the cloud there are
many advantages, and also many security challenges for the organizations regard-
ing cloud based data storage solutions. Cloud manufacturing is a novel manufactur-
ing architype developed from prevailing innovative manufacturing prototypes and
enterprise level information expertise under the provision of cloud computing, IoT,
virtualization technology and service oriented computing, and cutting-edge compu-
ting technologies and expertise. Manufacturing as a service is an advanced manu-
facturing architype developed from the existing cutting-edge manufacturing models
such as ASP, AM, NM and MGrid. This work is to study and discuss about security
and privacy issues and solutions in cloud based manufacturing.

Keywords: Cloud Computing, Cloud Security, Cloud Manufacturing, Cloud Privacy,

Internet of Things, Manufacturing as a Service, Encryption, Decryption

||
Sirisha Potluri, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI Foundation
for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana 501203, India,
[email protected]
Sachi Nandan Mohanty, Department of Computer Engineering, College of Engineering Pune,
Wellesley Rd, Shivajinagar, Pune, Maharashtra 411005, India, [email protected]
A. D. Sriram Kumar, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI
Foundation for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana
501203, India, [email protected]
D. Maheswari, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI Foundation
for Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana 501203, India,
[email protected]
B. Rahini, Department of CSE, IcfaiTech (Faculty of Science and Technology), ICFAI Foundation for
Higher Education, Donthanapally, Shankarpalli Road, Hyderabad, Telangana 501203, India,
[email protected]

https://doi.org/10.1515/9783110732573-009
172 | Sirisha Potluri, Sachi Nandan Mohanty, A. D. Sriram Kumar, D. Maheswari, B. Rahini

1 Introduction
Cloud computing is a great blessing for the very fast grow in the world of infor-
mation technology. Cloud computing is delivering the required and on demand
resources for their customers through the internet such as data storage, servers,
data bases, networking and software. It is a sophisticated architecture which stores
the data on remote servers and it can be accessed through the internet. Cloud based
storage allows us to store all the data in a safe and secure way and we can access it
whenever we required from anywhere just through the internet. Cloud computing
uses the payment option namely pay as you go basis means pay for what you use.
Many cloud providers in the marketplace namely Amazon web services, Google
cloud platform and Azure are effectively providing on demand resources to the
cloud consumers.
The computing resources are pooled together by the cloud service provider over
the infrastructure and the same can be provisioned to the users when they need
them. The cloud users or consumers will be charged and billed only for what they
have used or consumed from the cloud. Significant advantages of the cloud compu-
ting includes enhanced privacy, improved security, reduced costs, agility based
development, location transparency, extremely scalable, easy maintenance, device
independence, high performance, flexible and increase in productivity. To safe-
guard the data and information which is stored by the cloud service providers integ-
rity, availability and confidentiality is maintained by them.
Cloud manufacturing is also known as a novel manufacturing architype that is
developed from an existing and cutting-edge manufacturing models. The idea of
cloud manufacturing was originally recommended by a research group that was
headed by Prof. Bo Hu Li and Prof. Lin Zhang from China in the year 2009. After that
good amount of related research and discussions are being conducted and some
similar definitions to cloud manufacturing were introduced [1-5].

2 Manufacturing Cloud Provision Models

In the current market there are three types of services which are offered by the cloud
computing namely infrastructure as a service, software as a service and platform as
a service.
IaaS- Infrastructure as a service means that it provides business offers and ser-
vices based on pay as you go model. This can avoid investing in expensive storages,
servers, maintenance and gives users cloud based instead, so it delivers all those
services on the internet. The main advantage of the IaaS is that it provides highly
scalable services to the cloud users based on their demand. It ensures stability,
reliability and supportability, better security, quick response and rapid innovation.
 Cloud Manufacturing Service: A Secure and Protected Communication System | 173

SaaS- Software as a service makes use of all the software services through the
internet, they are like paid and use purpose. By using this SaaS, we do not need to
install any other kind of software applications on your computer. All these software
applications are available on the cloud and we can access them through the internet
when you require them.
PaaS- Platform as a service is one of the cloud computing service which uses vir-
tualization technology for the development platform for the cloud users and the
organizations. This platform is also include the memory, databases, computing,
storage, and also some other application development services. PaaS is mostly use-
ful for the environment for developing and the testing the applications. The main
advantages of this platform as a service is that it has no need to purchase hardware,
no need to spend more time, it will be speed up your applications, any person can
just login and use the applications from anywhere, and has high security for the
data being stored in it and no need to worry for the data loss [6-9].

3 Manufacturing Cloud Organization Models

Depending upon the deployment, cloud is divided into four models namely public
cloud, private cloud, hybrid cloud and community cloud.
Public Cloud- Public cloud provides all the resources to the cloud users and the
resources can be accessed by everyone through the internet. Examples for the public
cloud is Microsoft Azure, in public cloud you can share same storage and hardware
with other organisations. The key benefits of the public cloud are security, flexibil-
ity, controllability and cost effectiveness.
Private Cloud- Private cloud allows the resources only with the specific or se-
lected users instead of using them by all public users. It solves the security issues of
public cloud with the help of dedicated resource usage. The main advantages of this
private cloud are increase in security, efficient performance and competent capabil-
ity.
Hybrid Cloud- Hybrid cloud is a combination of both public cloud and the pri-
vate cloud. This type of cloud suits when an organisation uses the public cloud and
private for various requirements and demands. The advantages of the hybrid cloud
are flexibility, backup clouds, cost saving, security for the sensitive data.
Community Cloud- Community cloud shares the computing services to the
mainly targeted or some organisations. This type of cloud computing is also used for
the business organisations, research organisations and tenders and also heads of
trading firms. The main advantage of this community cloud are availability, reliabil-
ity, security and improved service provisioning [10-13].
Manufacturing cloud provision and organization models are represented in the
following fig. 1.
174 | Sirisha Potluri, Sachi Nandan Mohanty, A. D. Sriram Kumar, D. Maheswari, B. Rahini

Software as a service

Platform as a service

Infrastructure as a
service

Private cloud

Public cloud Hybrid cloud

Community cloud

Fig. 1: Manufacturing cloud provision and organization models

4 Cloud Manufacturing Paradigms

Cloud manufacturing is a parallel, thin, networked, interacted and distributed system
that includes an interlinked and integrated virtualized service pool of manufacturing
resources. This paradigm has the ability of intelligent management and controlling of
on demand use of services by providing solutions for all kinds of users that are includ-
ed in entire lifecycle of manufacturing [14-16]. The idea of cloud manufacturing im-
plies to a paradigm which includes the entire lifecycle of a product like production,
simulation, designing, test and maintenance. Applications of cloud based manufac-
turing are fabricated metal products, automobile manufacturing, housing, building
 Cloud Manufacturing Service: A Secure and Protected Communication System | 175

and furniture manufacturing, textiles, semiconductors and computers manufacturing,

transportation system manufacturing, distribution and warehousing, plastic and rub-
ber products, heavy manufacturing and electronics manufacturing are shown in fig. 2.

Fabricated
metal
products

Electronics Automobile

Housing,
Heavy building
machines and
Applications of
furniture
cloud
manufacturing

Plastic and
rubber Textiles
products

Trans- Semicon-
portation ductors and
system computers

Fig. 2: Applications of cloud based manufacturing

5 Benefits of Cloud Manufacturing

Benefits of cloud manufacturing are:

5.1 Scalability
Cloud centred software is a pay-as-you-go system, it means that during the demanding
times of the day/month/year the number of subscriptions that are used can be in-
creased. This helps to optimize the expenditures and keeps the spending predictable.
176 | Sirisha Potluri, Sachi Nandan Mohanty, A. D. Sriram Kumar, D. Maheswari, B. Rahini

5.2 Increased uptime

Cloud enables its users to continuously have the latest and state of the art version of
the software for its users. So, at whatever time the software is improved or updated
for better quality, everyone is benefitted by getting the updates immediately and
right away.

5.3 Standardization
Standardization improves the consistency of the resource utilization and usage by
streamlining the business practices and creation of standardization across the entire
operation.

5.4 Faster implementation

Due to faster implementation of cloud customer requirements, time and money are
saved.

5.5 Agility
Due to agility entire manufacturing process involves less operational cost by using
cloud manufacturing model.
Cloud computing system can be divided into three categories according to man-
ufacturing resources are follows:
– Service oriented and adapted input
– Operating and working platform of the manufacturing resource
– Service oriented and adapted output

There are three categories of key stake holders in cloud manufacturing system. They
are:
– Providers
– Operators
– Consumers

Providers will publish their resources to cloud manufacturing platform, the opera-
tors manage all aspects of platform that are relevant to the platform running and
operation, and the consumers requests the services from cloud platform.
A cloud based manufacturing platform consists of five layers as shown in fig. 3.
They are:
– Business planning
 Cloud Manufacturing Service: A Secure and Protected Communication System | 177

– Manufacturing operations management

– Automation and control
– Sensors and actuators
– Production process

Business planning: Define plant schedule

production, resource use, inventory and delivery

Manufacturing operations management: Manage

workflow control, record maintenance, and
process optimization

Automation and control: Real-time automation

and systems

Sensors and actuators: Sensing and actuating

the physical processes

Production process: The actual physical

processes

Fig. 3: Layer wise architecture of cloud based manufacturing platform

Cloud manufacturing is facing many challenges due to its integration with many
technologies. Cloud computing ensures high performance using semantic web,
internet of things and embedded systems. Several technical issues must be solved
such as cloud management engines, collaboration between cloud manufacturing
applications, knowledge base resource clouding and visualization and user inter-
face in cloud computing environments [17-18].
178 | Sirisha Potluri, Sachi Nandan Mohanty, A. D. Sriram Kumar, D. Maheswari, B. Rahini

6 Security and Privacy Aspects of Cloud

Manufacturing
Security in cloud computing is a set of control based integrated technologies & poli-
cies aimed to stick with the regular organizations, rules and protect/guard data
application and cloud technology infrastructure. Due to efficient resource sharing,
cloud security gives main problem to identity management, access control and pri-
vacy. The data in the cloud should be stored in an encrypted form. It processes the
security control in cloud and provides customer data security, compliance, and
privacy with necessary regulations.

6.1 Security challenges in cloud manufacturing

Cloud provides all its resources to many organisations in the subscription manner.
Though it has many advantages cloud needs to address many challenges as given
below.

6.1.1 Reliability

The consumers’ data are stored in third party’s server. They do not know where or
which location their data are being stored. They are not aware of what security
mechanisms provided to safeguard their data.

6.1.2 Privacy

All the cloud users store their data in the cloud and that is stored in some of the
physical machines in various places around the world. All the data is being trans-
ferred through the internet.

6.1.3 Data loss

Data loss is the one of the most security challenge in the cloud computing. If any
cloud user stores the important data in the cloud, we need to concern about the
security issues. If the data is lost through natural disasters or accidents or through
the DDoS it leads to disastrous to the organisation.
 Cloud Manufacturing Service: A Secure and Protected Communication System | 179

6.1.4 Bandwidth

When there is high demand of the resources or high volume of data is to be stored,
then high bandwidth is required.

6.1.5 Availability

At sometimes, when the huge amounts of resources are to be served at a time, then
the cloud becomes unable to give all the consumers’ demands.

6.1.6 Integrity

Cloud need to ensure the correctness and trustworthy of the consumers data. The
data should not be modified, maliciously fabricated, deliberately deleted or tam-
pered.

6.1.7 DoS

Denial of Service attack thrown by malicious users by selecting/choosing a victim

machine and by sending the bulk requests to the virtual machines at the same time,
then the services will not be made available to the reliable and true users.

6.1.8 Authentication

Every cloud user should be checked and authenticated before accessing the cloud
resources.

6.1.9 Access control

The cloud users or consumers need to be categorized and given permissions accord-
ing to the type of the cloud users.
Cloud service providers like Google cloud platform, Microsoft Azure, Amazon
web services are providing the best sources to the users to shift to the cloud compu-
ting. And users need not worry about their data security, because it’s being main-
tained by set of policies and protocols, virtual machines and data security as per the
rules and regulations.
180 | Sirisha Potluri, Sachi Nandan Mohanty, A. D. Sriram Kumar, D. Maheswari, B. Rahini

6.2 Significant factors effecting security in cloud system

Significant factors effecting security in cloud system are confidentiality, integrity,
and availability.

6.2.1 Confidentiality

Confidentiality is very important using which our data is not accessible to unauthor-
ized parties. It means that the sensitive data is accessible to only the authorized
persons. A failure in maintaining confidentiality means that unauthorized person
has the access to all the required sources provided by the provider which will be
caused due to intentional behaviour or by accident. This failure is known as confi-
dentiality failure or breach. Once the confidential data is leaked like bank details it
leads to high risks.
Requirements of confidentiality are divided into two categories namely data
confidentiality and virtualization confidentiality.

6.2.1.1 Data confidentiality

Cloud providers provide many online services to store the cloud user data. This may
sometimes result in potential loss of data security and confidentiality. When the
data is shared in a group using a distributed cloud storage service, the client should
get periodic update regarding the data changes by any user in the group. They are
many more interesting things by knowing the contents of the cloud user as well as
the cloud user’s access account and privilege information.

6.2.1.2 Virtualization confidentiality

In infrastructure as a service, cloud service provider hosts virtual machines when
there is a requirement from cloud users or customers. In cloud system, anyone with
the privileged access can read or change the deployed service. This means that the
total virtualization layer induces some security issues by considering all cloud secu-
rity related issues. Virtual machines allows elastic scaling, fault tolerance, hardware
pooling, easy maintenance, load balancing, and resource maintenance.

6.2.2 Integrity

Integrity is asset that promises the data or information is not been modified by any
third-party personnel who is not an authorized individual for such activity. Thus the
correctness and accuracy of an asset is with respect to its owner to ensure integrity.
Usually delete, edit, and append operations are believed by the change in integrity.
 Cloud Manufacturing Service: A Secure and Protected Communication System | 181

Consequently all the web related attacks are highly prevailing in cloud environment
that can alter or change the contents of the virtual machine metadata, databases,
user files/log files, and WSDL files.
Integrity requirements are divided into two categories namely data integrity and
virtualization integrity.

6.2.2.1 Data integrity

Cloud scheme allows decent number of data centric operations and tasks with very
huge data requirements where huge refers to terabytes and petabytes. Thus data
reliability challenges related with the PaaS (platform as a service), DaaS (data as a
service) etc. needs to be handled very carefully. Some of the cloud specific data
integrity issues are as follows:
– Data transfer/outsourcing
– SQL injection
– Cross scripting
– Metadata Spoofing attack
– Wrapping attack

6.2.2.2 Virtualization Integrity

As we have discussed in the earlier section, the virtualization layer it has of its own
security related challenges and issues which are not limited with privacy and securi-
ty but also integrity of the virtual machines. Some of the cloud specific virtualization
integrity issues are as follows:
– Cloud Service Provider administrators
– Virtual Machine replication
– Virtual Machine rollback
– Virtual Machine escape
– VM hopping

6.2.3 Availability

Availability is one of the significant aspects of security that has to be maintained

and preserved by the cloud service provider. Diverse business enterprises and initia-
tives who use the cloud based services to serve and work for their customers must
guarantee the highest availability of these services. Even a smallest amount of
downtime or stoppage of the service can result into a large financial loss which is
irredeemable.
This availability requirements are divided into two categories namely data
availability and virtualization availability
182 | Sirisha Potluri, Sachi Nandan Mohanty, A. D. Sriram Kumar, D. Maheswari, B. Rahini

6.2.3.1 Data availability

One of the significant issues affecting the cloud data availability is DDoS attack.
Different types of DDoS are recognised in the cloud environment are as follows.
– Denial of Service
– Indirect Denial of Service
– Natural disasters

6.2.3.2 Virtualization availability

Virtualization technology allows highest availability of resources and detects DDoS
attacks trough intrusion detection sensors. IBM smart cloud enterprise brought the
concept of virtual IP addresses for ensuring high availability of the cloud service
when there is an IP failover.

7 Comparison of Traditional/Conventional
Manufacturing and Cloud Manufacturing
Comparison of traditional/conventional manufacturing and cloud manufacturing
are given fig. 4 and fig. 5.
Traditional manufacturing suffers with the following limitations and are ad-
dressed by cloud manufacturing.
– Restricted and dedicated resources
– Static routing
– No interconnection and integration
– Independent control
– Isolated information
– Independent module setup
– Discrete work groups
– Lack of automation
– More manual work
 Cloud Manufacturing Service: A Secure and Protected Communication System | 183

No interconnection
and integration

Independent
Static routing
control

Restricted and Traditional/

Isolated
dedicated Conventional
information
resources manufacturing

Fig. 4: Limitations of traditional/conventional manufacturing

Self
Instant orgnization
interconnecti Bigdata
on

Dynamic
routing and Iot and
resource analytics
provisioning

Distributed
Cloud based
and dynamic Integration
manufacturing
resources

Fig. 5: Advantages of cloud manufacturing

184 | Sirisha Potluri, Sachi Nandan Mohanty, A. D. Sriram Kumar, D. Maheswari, B. Rahini

8 Secure Cloud Manufacturing Models

Table 1 is showing the analysis of secure cloud manufacturing models.

Tab. 1: Secure cloud manufacturing models

S.No Authors Secure cloud manufacturing model and its key practices

1 Carsten Ellwein, Oliver Riedel, The manufacturing process observe frequent changes
Olga Meyer, Daniel Schel [19] due latest current trends and technologies of digitaliza-
tion such as IoT, cloud computing and data analytics.
Cloud manufacturing model does influence product
production and manufacturing life cycle.
Key practices:
1. Digitalization
2. Industrial Internet of Things
3. Cloud manufacturing
4. Secure technology stack
5. Service-oriented architecture for modular
computing
6. Secured rent and produce (RnP) as a service
7. Integrated computer aided manufacturing
8. Computerized numerical control
2 Zhi Li, Ali Vatankhah Barenji, New evolving manufacturing prototypes such as cloud
George Q. Huang [20] based manufacturing, IoT facilitated manufacturing and
service-oriented architecture, have observed many
benefits to the manufacturing industry.
Key practices:
1. Distributed network for cloud manufacturing
2. Blackchain technology to ensure security
3. Secure data sharing
4. IoT enabled manufacturing and service-
oriented manufacturing
5. Distributed peer to peer network architecture
to preserve security and privacy in cloud man-
ufacturing (CMfg)
6. Blockchain based cloud manufacturing
(BCmfg) allows secure data sharing
7. Five layer integrated model is proposed to
achieve more security and scalability
3 Christian Esposito, Aniello Smart manufacturing upturns effectiveness and compe-
Castiglione, Ben Martini, Kim- tence through interconnection and collaboration among
Kwang Raymond Choo [21] organizations or among physical/logical/human re-
sources within a single company. Latest improvements
due to industry 4.0 supports collaboration, integration
and cooperation among the organization for better
 Cloud Manufacturing Service: A Secure and Protected Communication System | 185

S.No Authors Secure cloud manufacturing model and its key practices

resource sharing and usage

Key practices:
1. Collaborative networked manufacturing
2. Layered secure architecture of cloud manufac-
turing
3. Four phases of key management for secure
cloud manufacturing
4. Secure cloud service for defending against da-
ta breaches in cloud manufacturing
5. Networked manufacturing model
4 Haibo Yi [22] Cloud manufacturing has been acknowledged extensive
attentions everywhere the world. The expertise of cloud
manufacturing integrates services-oriented practices as
well as manufacturing practices based on cloud compu-
ting. With the assistance of the cloud computing plat-
forms, the manufacturing services are provided efficient
secure solutions to make it as a most popular choice.
Key practices:
1. Post-quantum asymmetric-key encryption
scheme for data security
2. Post-quantum public-key signature genera-
tion for authentication security
3. Post-quantum secure communication system
for secure and efficient cloud manufacturing
4. Safeguarding cloud manufacturing communi-
cation
5. Secure and privacy preserving encryption and
decryption system

5 Florin Anton, Theodor Borangiu, Cloud systems ensuring more production trough indus-
Silviu Răileanu, Silvia Anton, try 4.0 solutions, integrating shop-floor processes and
Nick Ivănescu, Iulia Iacob [23] managing higher-level enterprise modules.
Key practices:
1. Cloud based manufacturing system integra-
tion
2. Remote access and control to manufacturing
system
3. VM based custom configured cloud
186 | Sirisha Potluri, Sachi Nandan Mohanty, A. D. Sriram Kumar, D. Maheswari, B. Rahini

9 Cloud manufacturing Life Cycle: An Efficient

Integrated Approach
IoT, data analytics, big data, self-organization, and efficient resource utilization is
achieved through efficient integration of cloud manufacturing as shown in fig. 6.

Raw materials

Production
process

Delivery

Customer
utilization

Disposal and
recycling

Fig. 6: Cloud manufacturing- Integrated approach

 Cloud Manufacturing Service: A Secure and Protected Communication System | 187

10 Proposed Secured Cloud Manufacturing Model

Architecture
The proposed architecture is illustrated in fig. 7. The proposed model has asymmet-
ric-key encryption scheme for data security, public-key signature generation for
authentication security and secure communication sys-tem for secure and efficient
cloud manufacturing to safeguard cloud manufacturing system communication and
data security.
Different keys are used to encrypt and decrypt the data at sender and receiver
side. With improved encryption mechanism plain text is converted into cipher text.
Using efficient decryption mechanism cipher text is converted in to plain text.

Plaintext

Encrypt
Sender

Different keys are used Ciphertext

to encrypt and decrypt
the message

Recipient
Decrypt

Plaintext

Fig. 7: Architecture of secure cloud manufacturing model

The following code shows secure data encryption using python platform.
188 | Sirisha Potluri, Sachi Nandan Mohanty, A. D. Sriram Kumar, D. Maheswari, B. Rahini

Listing 1: Secure data encryption: Python code

# Key encapsulation
from pprint import pprint
import oqs
km = oqs.get_enabled_KEM_mechanisms()
pprint(km, compact="True")
kmmech = "DEFAULT"
with oqs.KeyEncapsulation(kmmech) as clnt:
with oqs.KeyEncapsulation(kmmech) as srv:
pprint(clnt.details)
p_k = clnt.generate_keypair()
c_txt, sh_s_s = srv.encap_secret(public_key)
sh_s_c = clnt.decap_secret(ciphertext)
print(sh_s_c)
# Random number generation
import platform, oqs.rand as oqsrand
ornd.randombytes_nist_kat_init_256bit(bytes(entropy_seed))
#Check system settings and switch the algorithm
# Key signature generation
s_sig = oqs.get_enabled_sig_mechanisms()
pprint(s_sig, compact="True")
msg = "This is message from key generation message".encode()
s_a = "DEFAULT"
with oqs.Signature(s_a) as i_sig:
with oqs.Signature(sigalg) as i_ver:
pprint(i_sig.details)
s_p_k = i_sig.generate_keypair()
sgn = i_sig.sign(msg)
is_v = i_ver.verify(msg, sgn, s_p_k)
#Post quantum key encryption
pip install pynewhope
from pynewhope import newhope
p_key, message1 = newhope.keygen()
s_key, message2 = newhope.sharedB(message1)
sh_key = newhope.sharedA(messsage2, p_key)
if sh_key == s_key:
print("Success")
else:
print("Fail")
 Cloud Manufacturing Service: A Secure and Protected Communication System | 189

11 Conclusion
Mobile and instant communication for machines is increasing the importance of the
manufacturing industry. Smooth movement of operations, improving production,
automation of production life cycle, cloud based security, IoT based services and
offerings, service oriented computing paradigms, advanced encryption and decryp-
tion mechanisms, faster growth and platform safety are driving cloud manufactur-
ing to a next high. Cloud based manufacturing ensures reduced costs, on demand
delivery, efficient data management and efficient management of human resources
through competent practices and protocols.

12 References
[1] Tao F, Zhang L, Venkatesh V, Luo Y, Cheng Y, Cloud manufacturing: a computing and service-
oriented manufacturing model, Proceedings of the Institution of Mechanical Engineers, Part B:
Journal of Engineering Manufacture, 2011, 225(10), 1969-1976, doi:10.1177/0954405411405575
[2] Li B, Zhang L, Ren L, Chai X, Tao F, Luo Y, Wang Y, Yin C, Huang G, Zhao X, Further discussion on
cloud manufacturing, Comput. Integr. Manuf. Syst, 2011, 17(3), pp. 449–457
[3] Ren L, Cui J, Li N, Wu Q, Ma C, Teng D, Zhang L, Cloud-based intelligent user interface for cloud
manufacturing: model, technology, and application, ASME, J. Manuf. Sci. Eng, August 2015,
137(4), 040910, https://doi.org/10.1115/1.4030332
[4] Wu D, Thames J, Rosen W, Schaefer D, Towards a cloud-based design and manufacturing para-
digm: looking backward, looking forward, Proceedings of the ASME 2012 International Design
Engineering Technical Conferences and Computers and Information in Engineering Conference,
Volume 2: 32nd Computers and Information in Engineering Conference, Parts A and B, Chicago,
Illinois, USA, 2012, pp. 315-328, https://doi.org/10.1115/DETC2012-70780
[5] Wu D, Greer J, Rosen W, Schaefer D, Cloud manufacturing: strategic vision and state-of-the-art,
J. Manuf. Syst, 2013, 32(4), pp. 564–579, 10.1016/j.jmsy.2013.04.008
[6] Xu X, From cloud computing to cloud manufacturing, Rob. Comput. Integr. Manuf, 2012, 28(1),
pp. 75–86, 10.1016/j.rcim.2011.07.002
[7] Dazhong W, Matthew J, David W, Dirk S, Cloud manufacturing: strategic vision and state-of-the-
art, Journal of Manufacturing Systems, 2013, Volume 32, Issue 4, Pages 564-579, ISSN 0278-
6125, https://doi.org/10.1016/j.jmsy.2013.04.008
[8] Lin Z, Yongliang L, Fei T, Bo H, Lei R, Xuesong Z, Hua G, Ying C, Anrui H, Yongkui L, Cloud manu-
facturing: a new manufacturing paradigm, Enterprise Information Systems, 2014, 8:2, 167-187,
DOI: 10.1080/17517575.2012.683812
[9] Lei R, Lin Z, Fei T, Chun Z, Xudong C, Xinpei Z, Cloud manufacturing: from concept to practice,
Enterprise Information Systems, 2015, 9:2, 186-209, DOI: 10.1080/17517575.2013.839055
[10] Lihui W, Machine availability monitoring and machining process planning towards cloud manu-
facturing, CIRP Journal of Manufacturing Science and Technology, 2013, Volume 6, Issue 4,
Pages 263-273, ISSN 1755-5817, https://doi.org/10.1016/j.cirpj.2013.07.001
[11] Petri H, Duy P, Yuqiuge H, Cloud manufacturing – scheduling as a service for sheet metal man-
ufacturing, Computers & Operations Research, 2019, Volume 110, Pages 208-219, ISSN 0305-
0548, https://doi.org/10.1016/j.cor.2018.06.002
190 | Sirisha Potluri, Sachi Nandan Mohanty, A. D. Sriram Kumar, D. Maheswari, B. Rahini

[12] Potluri S, Primary methods to address the data security problems in cloud computing, The IUP
Journal of Computer Sciences, 2016, Vol. X, Nos. 1 & 2, pp. 18-24, 2016
[13] Avinash M, Potluri S, A study on technologies in cloud-based design and manufacturing, Inter-
national Journal of Mechanical and Production Engineering Research and Development
(IJMPERD), 2018, Volume 8, Issue 6, PP: 187-192, ISSN(P): 2249-6890; ISSN(E): 2249-8001
[14] Potluri S, Subbarao K, Quality of service-based cloud models in manufacturing process auto-
mation, Innovations in Computer Science and Engineering. Lecture Notes in Networks and Sys-
tems, 2019, vol 32, PP: 231-240, Springer, Singapore. https://doi.org/10.1007/978-981-10-
8201-6_26
[15] Zhang Y, Zhang G, Liu Y, et al., Research on services encapsulation and virtualization access
model of machine for cloud manufacturing, 2017, J Intell Manuf, 28, 1109–1123,
https://doi.org/10.1007/s10845-015-1064-2
[14] Wang X, Mohammad G, Lihui W, Manufacturing System on the Cloud: a case study on cloud-
based process planning, Procedia CIRP, 2017, Volume 63, Pages 39-45, ISSN 2212-8271,
https://doi.org/10.1016/j.procir.2017.03.103
[15] Yingfeng Z, Geng Z, Ting Q, Yang L, Ray Y, Analytical target cascading for optimal configuration
of cloud manufacturing services, Journal of Cleaner Production, 2017, Volume 151, Pages 330-
343, ISSN 0959-6526, https://doi.org/10.1016/j.jclepro.2017.03.027
[16] Yongkui L, Lihui W, Vincent W, Xun X, Lin Z, Scheduling in cloud manufacturing: state-of-the-art
and research challenges, International Journal of Production Research, 2019, 57:15-16, 4854-
4879, DOI: 10.1080/00207543.2018.1449978
[17] Ellwein C, Riedel O, Meyer O, Schel D, Rent ‘n’ Produce: a secure cloud manufacturing platform
for small and medium enterprises, 2018 IEEE International Conference on Engineering, Tech-
nology and Innovation (ICE/ITMC), 2018, pp. 1-6, doi: 10.1109/ICE.2018.8436332
[18] Zhi L, Vatankhah B, George Q, Toward a blockchain cloud manufacturing system as a peer to
peer distributed network platform, Robotics and Computer-Integrated Manufacturing,2018,
Volume 54, Pages 133-144, ISSN 0736-5845, https://doi.org/10.1016/j.rcim.2018.05.011
[19] Esposito C, Castiglione A, Martini B, Choo K, Cloud manufacturing: security, privacy, and fo-
rensic concerns, IEEE Cloud Computing, 2016, vol. 3, no. 4, pp. 16-22, doi:
10.1109/MCC.2016.79
[20] Yi H, A post-quantum secure communication system for cloud manufacturing safety, J Intell
Manuf , 2021, 32, 679–688, https://doi.org/10.1007/s10845-020-01682-y
[21] Anton F, Borangiu T, Răileanu S, Anton S, Ivănescu N, Iacob I, Secure sharing of robot and
manufacturing resources in the cloud for research and development, Berns K, Görges D, Ad-
vances in Service and Industrial Robotics, RAAD 2019, Advances in Intelligent Systems and
Computing, 2019, vol 980, Springer, Cham. https://doi.org/10.1007/978-3-030-19648-6_61
Index
amazon web service 85 hybrid cloud 53
artificial intelligence 4 hypervisor 54
asymmetric key encryption scheme 187
autonomous or self governing vehicle 152 information security 44
availability 68 infrastructure as a service 67
integrity 67, 180
basic safety message 147 intelligent transportation systems 162
blockchain as a service 31 internet enabled sensors 92
blockchain structure 29 internet of things 84
blockchain technology 23 interoperability 129
IoT based testing and information processing
caesar cipher 17 102
capital expense 105 IoT compressed IoT devices 103
ciphertext 11 IoT cyber security framework 86
client server computing 22 IoT gadget connectivity 88
cloud based computing 24
cloud based management 93 landing page optimization 112
cloud centred software 175 laser illuminated detection and ranging 153
cloud computing 2
cloud governance 108 machine learning 126
cloud manufacturing 174 manufacturing cloud provision and organization
cloud protection 33 173
cloud service providers 179 marketing strategy 107
cloud shell 137 mobile and instant communication 189
cloud system security 68
community cloud 53 173 network security and privacy 87
computational techniques 132
confidentiality 67 on premise data storage 128
consensus mechanism 28 operating expenditure 105
conversion rate optimization 112
cryptography 89 P2P cloud computing 73
customer relationship management 113 pay-per-click strategy 113
cyber attacks 15 platform as a service 67
private cloud 53
data integrity 3 proof of burn 29
data loss 178 proof of stake 29
data privacy 75 proof of work 28
data security 77 pseudonyms coupled with public key
data usage 3 infrastructure 161
dedicated short range communication 146 public cloud 53
denial of service 42
denoising autoencoder 14 quality of service 147
distributed denial of service 42
RACE Framework 109
google cloud machine learning engine 132 radio frequency identification 84
group signatures 161 reliability 129

https://doi.org/10.1515/9783110732573-010
192 | Index

resilience 34 software as a service 66

resource management 130 software defined networks 43
return on investment 118
reverse cipher 17 third party based encryption scheme 3
traditional conventional manufacturing 182
safe road trains for the environment 153 trust based schemes 161
scalability 129
security information and event management vehicular ad hoc networks 146
system 137 virtualization 54
service level agreement 43