262

ISSN: 2057-5688
A CUTTING-EDGE HEALTH MONITORING SYSTEM
BASED ON IOT AND CLOUD COMPUTING THAT USES
MACHINE LEARNING.

Srinivasa Rao Kadari, Research Scholar, Department of Computer

science , Radha Govind University, Ramgarh,Jharkhand.
Dr. Sanjay Kumar ,Assistant Professor ,Supervisor, Department of
Computer Science ,Radha Govind University, Ramgarh,Jharkhand.

Abstract:
The Internet of Things (IoT) has recently become one of the IT industry's most well-
liked developing technologies. IoT is defined as a network of connected, intelligent
physical objects. Through the use of wired or wireless networks, sensors are
integrated into physically linked objects and communicate with one another. The
interconnectedness of devices, smart, dynamic nature, sensing, enormous size,
heterogeneity, and security are the main characteristics of IoT. A consumer may
access a variety of cloud services across a network, including database, application,
and storage. The Internet of Things (IoT) provides a wide variety of field applications
for ongoing monitoring in many different fields, with health care being one of them.
The Internet of Things (IoT) has become entrenched in industries that handle
enormous amounts of data, particularly with the introduction of IoT-cloud-based
devices. One of the newest uses for the IoT-Cloud is the healthcare system. Numerous
studies are conducted to protect the confidentiality of patient data. The security of
data and computational overheads continue to be the key problems with the IoT-based
cloud-based health system. Another challenging area of health systems is the ability to
predict illness using patient data from the IoT device.

Keywords: IoT, Cloud, Health, Data, Security

Volume XIII Issue I1 2021 MARCH http://ijte.uk/ 262

 263

ISSN: 2057-5688

1 INTRODUCTION: On-demand services: Using the cloud,

There are scalable resources available consumers may get services for as long
in several subscription plans on the as they need them and as per their
cloud. In the end, you'll only be expectations. According on how much
charged only for resources are actually they used at that specific time period,
used. This helps you deal with demand the users get charged. Users get these
increases without having to invest in services based on demand and
long-term hardware. specifications.

As an example, Netflix makes use of Broad network access: Users may

cloud computing's advantages, which it access cloud computing services using
provides. The company experiences thick or thin clients on a heterogeneous
significant increases in server usage platform by exploiting the underlying
during periods of peak traffic due to its network infrastructure in accordance
streaming on demand service. Moving with conventional processes. Users
from its internal data centers to the may connect, for instance, through
cloud allowed the company to increase laptops, smartphones, desktop
its customer base exponentially computers, or tablets.
without the need to invest cash on the
setup and maintenance of costly Resource Pooling: The cloud's
devices. resource sharing function is
accomplished by generating virtual
ESSENTIAL FEATURES OF instances of computing resources,
CLOUD COMPUTING which are then made accessible to
Cloud computing, or just "Clouds," has users across the network. Network,
a number of distinctive characteristics. storage, servers, apps, and web
Six crucial features are covered out of services are examples of computing or
them. storage resources that are kept as a
pool of resources, and many users are

Volume XIII Issue I1 2021 MARCH http://ijte.uk/ 263

 264

ISSN: 2057-5688
permitted access to those resources for an affordable price by metering and
a certain period of time as needed. billing consumers for the specific
period of time they use a certain
Rapid elasticity is the capacity of cloud service via a specific resource.
computing to manage utilisation by
scaling up and down of resources 2 LITREATURE SURVEY:
during peak demands and off-peak In the Indian healthcare industry,
workloads. By making resources electronic health records (EHRs) are a
available as needed, this feature recent development. According to
enables our apps to operate without a studies from the Ministry of
hitch while adjusting to shifting Electronics and Information
requirements. Technology (Sunil Kumar, 2016),
EHR adoption is crucial for the
Measured service: Depending on the development of smart health systems.
kind of service (for instance, storage, For effective HR management,
CPU, network bandwidth, servers, user effective strategies must be used. The
accounts, etc.) offered to the user, four most crucial aspects must be
cloud systems intuitively manage and addressed. These include the ICT
enhance resource allocation and its infrastructure, rules and regulations,
utilisation using metering mechanisms standards and interoperability, and
at the relevant layer abstraction. R&D with training. The administration
Resource utilisation is overseen for of HRs includes a variety of tasks,
monitoring, administered for such as sharing and exchanging safe
controlling, and reported, giving cloud information.
clients transparency. Due to the fact
that cloud services are metered, Both clinical and non-clinical practises
consumers are only charged for the may benefit from the archiving of
services they really use. The business healthcare data. In 2012, Shreekant
model is known as "Pay-per-use" or Iyengar and colleagues performed a
"Pay-as-you-go," and it allows study on the availability of primary
customers to access cloud services at healthcare for Indians living in rural

Volume XIII Issue I1 2021 MARCH http://ijte.uk/ 264

 265

ISSN: 2057-5688
areas. Data from the states of Madhya disease in India's rural areas. The
Pradesh, Uttar Pradesh, Rajasthan, SMART health India initiative, which
Karnataka, Andhra Pradesh, and Tamil includes 18 PHCs in rural areas, goes
Nadu have been gathered for this by that name. Home CVD risk
survey's clinical and non-clinical data assessment data, clinical decision
sections. To maintain democratic support data, and patient monitoring
accountability, Shirin Madon et al. data were gathered.
(2010) also discussed the significance
of decentralisation in health A summary of several issues of
information systems. The biggest managing EHRs is provided by Haas et
democratic nation in the world is India. al. (2011). Data gathering is done
The country has its own laws to uphold using the centralised health record
the transparency of all services management system. This framework
provided to its residents. will increase accessibility and
Accountability must be maintained in comprehensiveness. The upkeep of HR
health information systems as well. records in a database presents
difficulties with regard to security,
The information needed to create the privacy, data concealment, and
tool for bettering mental health has accessibility to authorised users. A
been collected by Rahul Shidhaye et al. different data service may be utilised
in 2019. For the purpose of designing to connect users, such as patients,
the tool's several stages, the data from medical systems, external storage, and
the year 2013 was analysed. The study physicians, depending on the demand.
was conducted in Madhya Pradesh's
Sehore district. A large database of the The statistics on reproductive
patients over a six-year period is healthcare in India, according to
created using the information gathered Amlan Majumder & Upadhyay (2004),
from numerous testing. According to may be used to predict future social
Peris D et al. (2019), there are several and economic issues. Data collecting
ways to gather information on the on service use, accessibility, and
increased risk of cardiovascular availability, family characteristics,

Volume XIII Issue I1 2021 MARCH http://ijte.uk/ 265

 266

ISSN: 2057-5688
social structure, and level of care are attempted, where little or no research is
crucial in the area of reproductive done, some revelations are not
healthcare. Keeping clinical records is discovered, and evaluation may give a
beneficial for comparing diagnosis. more flexible method of working or
Virostko et al. (2016) looked at the bearing to meet certain examination
utility of keeping electronic medical areas with the newest resources.
records to assess patients with type 1
diabetes' pancreas size. Instead of attempting to foster another
clinical idea security system or
3 PROPOSED METHOD investigate the possibility of conveying
one, this examination used a
In this exploratory analysis, "observing positivistic assessment approach that
flaws in irrelevant or prior assessments adequately analysed prior enthusiastic
is endeavoured." Early evaluation data while surveying the security of
findings suggest that experts may use individual information in current IoT-
this form of assessment to somewhat, based clinical idea structures. The
but not significantly, alter the course of maker's positivistic evaluation
their request efforts. It isn't even close approach was an effective solution to
to being as precise as the tested testing the appraisal problem for a number of
methods. The focus on action and its reasons. Such a system recognised the
findings in the specific situation that manner that IoT-based clinical
was previously discussed will probably advantage structures exist and
be clarified by Last's investigation. In dissected their strength, benefits,
any event, by describing how research limitations, and potential starting with
activity may be shifted about, it adds to the onion approach for investigation.
the systematic evaluation. The To guide the testing, information from
assessment strategy is seen as a a variety of sources was used. Due to
thorough technique for responding in the expert's thorough analysis of the
this line of thinking. My exploration situation, the examiner had the
review is set up with the test centre opportunity to get a thorough
since new master hits are often understanding of the present situation,

Volume XIII Issue I1 2021 MARCH http://ijte.uk/ 266

 267

ISSN: 2057-5688
as well as if and how it may be most common types. Recently,
resolved. quantitative data were shown and
divided into inferential, test, and
A strategy for dispersing vulnerability stimulatory tasks. to discuss the
by collecting, examining, and grabbing informational gathering technique used
each cycle one at a time is included in to compile the affiliations of residents.
an exploratory approach. As a follow- The broad audience is used as an
up, researchers should understand why example to imply that everyone will
people need to reflect on or examine find the review's tangential findings to
the reality of the outside world. A few be instructive. Various components are
successful frameworks, designs, and used to assess the primary association
innovations were thoroughly evaluated in the evaluation environment. The
using data from a preceding research creation of breathtakingly skewed or
as well as the strong evidence that replicated environmental realities is a
information on patients' clinical component of entertainment. It
considerations might be reached. conveys the impression that an animal
Because of advancements in is growing and evolving in a stable
innovation, there are challenges and environment. Innovative and socially
solutions for developing the conscious application: In this particular
exploration structure further. case, "redirection" refers to "the
Subjective, quantitative, and integrated example of a genuine model, tending
assessment techniques are three of the to the plan of a striking activity."

Volume XIII Issue I1 2021 MARCH http://ijte.uk/ 267

 268

ISSN: 2057-5688

Figure 3.1: Research Methodology for Proposed Research

This study was created using descriptions. The data was then
information from sources that were at evaluated to see if it was consistent.
least eight to ten years old. Monster The most important information was
hotspots were found, each piece of kept, along with the presumptions that
evidence was carefully examined, and were based on it.
significant efforts were made to
decipher convoluted language and

Volume XIII Issue I1 2021 MARCH http://ijte.uk/ 268

 269

ISSN: 2057-5688
It may be challenging to maintain years in order to eradicate outdated and
validity in research measures that obsolete data. Material that had been
incorporate data from several sources. on display but hadn't been utilised for a
Other sources included research papers, while was thrown away.
electronic databases, journal articles,
data collection methods, and HOMOLOGUE ENCRYPTION
connection complaints. High impact The technique referred to as
journal articles were picked from all homomorphic encryption (HE) (4)
around the globe and either refereed or permits sophisticated computations to
peer-reviewed to ensure validity. be carried out with encrypted data,
Representatives from renowned without compromising the security of
universities, social groups for encryption. In math, HE is a reference
educators, and mechanical to the procedure of turning an existing
organisations were contacted for dataset to another, with the same
further information. The analysis of connections maintained between the
several data sources produced valuable elements in the two sets. The Greek
material, around thirty of which were terms that refer to "same structure" are
pertinent evaluation reports. The where this phrase comes from. This is
delivery of clear contents at otherwise why logic-based calculations
happy occasions, such as exam papers, performed on encrypted or decoded
summaries, extended developments, data employ the same schema as the
speeches, well-known copies, and data used in this homomorphic
other materials, was also considered as encryption system.
a reliable source of important
information. The retention period for 4 EXPERIMENT AND RESULTS
these systems was extended to eight

Table 4.1 Access Control Policy Generation

Volume XIII Issue I1 2021 MARCH http://ijte.uk/ 269

 270
Generation Key
An algorithm can be used to create
keys in cryptography which is also
known as key creation. Keys created
by algorithm are typically used for
encryption of data and decryption. A
random choice of an integer "k" within
between [1, N-1] gives an encrypted
private key (keypr).The Generator
function Gc as well as the public key
(keypr) can be employed to establish
that the private key (keypu).
Key generation in algorithm I
Generator function (Gc) as well as a
random numbers (k) serve as two
inputs.
Public Key (keypu) as well as Private
Key (keypr) output
1. An random number will be selected
from the number interval [1,n-1] in
order to determine the private secret.
Volume XIII Issue I1 2021
ISSN: 2057-5688
Keypr (private key) = K
Determine the your public key (keypu)
is keypr*Gc the second step.
Encoding of Messages
Converting messages into elliptic
curve points is known as message
encryption. The ElGamal Curve
(EGEC) techniques are unable to be
used to decrypt or encode the input text
information; they are able to only
decrypt and encrypt points along the
curvature. Think of an input message
to be encrypted with the specified
methods. The message input is divided
into blocks of fixed size that are just
one character. Each ASCII number is
immediately converted to the Elliptic
Curve point once each word in the text
is transformed into the ASCII value.
Algorithm II: Encoding of Messages
input: message in plain text
Elliptic Curve points as output
MARCH http://ijte.uk/ 270

To begin, you must divide messages
into blocks of fixed size that are
composed of one word.
The second step is translating every
text message character into
appropriate ASCII value.
3. Elliptic Curve points are instantly
transformed to ASCII value.
Encryption using EGEC
An approach to convert plain text to
cyphertext is known as encryption. The
Elgamal-Elliptic curvature (EGEC)
Homomorphic encryption system that
is suggested in this work can secure
plain text messages as well as increase
Table 4.2 Comparison of Encryption and Decryption time
Volume XIII Issue I1 2021
271
ISSN: 2057-5688
security at the lowest cost. The data
input to this method is the point on the
curve (Pm) and keys for private and
public (keypr) as well as keys for the
public (keypu) to a particular chunk of
the text. The random 4-bit number "rn
generated by the generator function,
'Gc can be used to generate the
CipherText point "CT1'. This is then
done by using the curve point "Pm" for
a certain message block, the random
number "rn," and thepublic key
(keypu), CipherText point "CT2" is
created. A collection of cypher text
points is subjected to homomorphic
computations using either the additive
property or the multiplicative property.
Finally, the cloud is used to store the
encrypted point Ep= (CT1,CT2).
MARCH http://ijte.uk/ 271
 272

ISSN: 2057-5688

Figure 4.1 Comparison of Encryption Time

Decryption Time 1024KB,1536KB, 2048KB and

The period of time that is required for
the algorithm to convert the encrypted
text into plain text is termed as
decryption time. Milliseconds could be
used as a way of describing the
duration of decryption. Figure 4.5
shows the time taken to decrypt for the
new EGEC method with the existing
methods. Based on Table 4.7 it is
apparent that current methods MORE
use 220ms, 280ms and 385ms 498ms,
590ms and 220ms to encrypt input
sizes that are 512KB, 102KB, 1536KB,
2048KB and 2560KB. On the other
hand, PORE is 210ms long, 279ms
long 3,78ms, 470ms and 575ms. The
time to decrypt inputs of 512KB,
2560KB when using the suggested
EGEC homomorphic approach include
180ms and 250ms. 350ms, 400ms, and
495ms. In comparison to the current
methods the proposed EGEC
homomorphic technique offers a
quicker decryption speed..
5 Conclusion
Partial homomorphic encryption was
used in the creation of a secure EHR
storage system for patient records. The
PHE-BGN Cryptosystem is used to
store sensitive patient data in the cloud
while maintaining its confidentiality
and integrity. Using the CRT and
BSGS algorithms, the performance of

Volume XIII Issue I1 2021 MARCH http://ijte.uk/ 272

 273

ISSN: 2057-5688
the BGN crypto system is enhanced. utilization along with encryption
The system's main benefits are that 1) throughput as well as encryption
EHRs are safely stored in the cloud throughput are employed to evaluate
and 2) EHRs are safe even while being the efficacy for the proposed ElGamal
transported since BGN Homomorphic Elliptic curve homomorphic encryption
encryption permits aggregate actions method. The system is compared to
on the encrypted data. According to this proposed EGEC homomorphic
experimental findings, the BGN encryption method as well as current
cryptosystem is more effective than the systems like PORE and PORE.
traditional Paillier Algorithm. Security 6 REFERENCE:
of cloud-based data can be assured by 1. Abbas, A & Khan, SU 2014, ‗A
ElGamal Elliptic curve homomorphic review on the state-of-the-art
encryption which has been proposed. privacypreserving approaches in the e-
Six different processes are part of the Health clouds‘, IEEE Journal of
suggested EGEC homomorphic Biomedical and Health Informatics,
encryption. These include key vol.18(4), pp. 1431–1441.
generation, message encryption, EGEC 2. Abhijit V Banerjee, Rachel
encryption, EGEC decryption and Glennerster & Esther Duflo 2008,
message decoding. When an end user ‗Putting a Band-Aid on a Corpse:
wants access to the data in cloud Incentives for Nurses in the Indian
storage it is the CSP confirms the Public Health Care System‘, Journal of
access rights of that user. By using the European Economic Association,
EGEC homomorphic encryption owner vol. 6(2-3), pp. 487-500.
of the data secures the data. For 3. Acar, A, Aksu, H, Uluagac, AS &
encrypted data it is homomorphic Conti, ε 2017, ‗A Survey on
process performed. In order to recover Homomorphic Encryption Schemes:
the original message to retrieve the Theory and Implementation‘, pp. 1–35.
original message, the EGEC https://doi.org/10.1145/0000000.00000
encryption technique is employed. 00.
Time to execute, encryption time and 4. Alabdulatif, I, Khalil, X, Yi &
decryption times and memory Guizani, ε 2019, ‗Secure Edge of

Volume XIII Issue I1 2021 MARCH http://ijte.uk/ 273

 274

ISSN: 2057-5688
Things for Smart Healthcare based cohort study‘ Diabetes Care,
Surveillance Framework,‘ in IEEE vol.33, no. 12, pp. 2580-2585.
Access, vol. 7, pp. 31010-31021. 10. Aslett, LJM, Esperança, PM &
5. Alex Roehrs, Cristiano André da Holmes, CC 2015, ‗A review of
Costa & Rodrigo da Rosa Righi 2017, homomorphic encryption and software
‗OmniPHR: A distributed architecture tools for encrypted statistical machine
model to integrate personal health learning‘, pp. 1–21. Retrieved from
records‘, Journal of Biomedical http://arxiv.org/abs/ 1508. 06574.
Informatics, vol. 71, pp. 70-81. 11. Ateniese, G, Fu, K, Green, M &
6. Alyami, ε 2017, ‗εanaging personal Hohenberger, S 2006, ‗Improved
health records using meta-data and proxy re-encryption schemes with
cloud storage.‘ IEEE/ACIS 16th applications to secure distributed
International Conference on Computer storage‘, ACM Transactions on
and Information Science (ICIS) pp. Information and System Security,
265-271. vol.9(1), pp. 1–30.
7. Amlan εajumder, V & Upadhyay 12. Awasthi, P, Mittal, S, Mukherjee, S
2004, ‗n analysis of the primary health & Limbasiya, T 2019, ‗A Protected
care system in India with focus on Cloud Computation Algorithm Using
reproductive health Care services‘, Homomorphic Encryption for
ArthaBeekshan, vol.12(4), pp. 29-38. Preserving Data Integrity‘, In: Sa P,
8. Anthony Wellever, Gerald Hill & Bakshi S, Hatzilygeroudis I, Sahoo M.
εichelle Casey 1998, ‗Commentary: M (eds) Recent Findings in Intelligent
dicaid Reform Issues Affecting the Computing Techniques. Advances in
Indian Health Care System‘, Public Intelligent Systems and Computing,
Health Policy Forum, vol.88(2), pp. vol. 707. Springer, Singapore
193-195. 13. Bahga, A & εadisetti, VK 2013, ‗A
9. Antonio Gonzalez-Perez, Raymond cloud-based approach for interoperable
G Schlienger & Luis A García electronic health records (EHRs)‘,
Rodríguez 2010, ‗Acute Pancreatitis in IEEE J Biomed Health Inform.
Association With Type 2 Diabetes and vol.17(5), pp. 894-906.
Antidiabetic Drugs, A population-

Volume XIII Issue I1 2021 MARCH http://ijte.uk/ 274

 275

ISSN: 2057-5688
14. Bajpai, Nirupam & Goyal, 20. Bondale, N, Kimbahune, S &
Sangeeta 2004, ;Primary Health Care Pande, A 2013, ‗mHEAδTHPHC: An
in India: Coverage and Quality Issues‘, ICT Tool for Primary Healthcare in
CGSD Working Paper, vol.15, pp. 1- India‘, IEEE Technology and Society
39. Magazine, FALL 2013, pp. 31-38.
15. Basu, S 2012, ‗Fusion: εanaging 21. Camilla L Cunha, Alexandre R
Healthcare Records at Cloud Scale,‘ in Torres & Aderval S Luna 2020,
Computer, vol. 45, no. 11, pp. 42-49. ‗εultivariate regression models obtaine
16. Bertrand, V, Smokvina, E, Masson, d from near-infrared spectros copy data
E & Bruel, H 2019, ‗Severe acute for prediction of the physical
pancreatitis in a child with properties of biodiesel and its blends‘,
phenylketonuria,‘ Arch. Pédiatrie, vol. Fuel, vol. 261, p.116344.
26, no. 2, pp. 2018–2020. 22. Chen, ε, Hao, Y, Hwang, K, Wang,
17. Bitewulign Kassa Mekonnen, δ & Wang, δ 2017, ‗Disease
Webb Yang, Tung-Han Hsieh, Prediction by Machine Learning Over
ShienKueiLiaw, Fu-Liang Yang, Big Data From Healthcare
18. Bitewulign Kassa Mekonnen, Communities‘, in IEEE Access, vol. 5,
Webb Yang, Tung-Han Hsieh, pp. 8869-8879.
ShienKueiLiaw & Fu-δiang Yang 2020,
‗Accurate prediction of glucose
concentration and identification of
major contributing features from
hardly distinguishable near-infrared
spectroscopy,‘ Biomedical Signal
Processing and Control, vol.59.
19. Bocu, R & Costache, C 2018, ‗A
homomorphic encryption-based system
for securely managing personal health
metrics data‘, IBε Journal of Research
and Development, vol. 62(1), pp.1:1-
1:10.

Volume XIII Issue I1 2021 MARCH http://ijte.uk/ 275