Iot Enabled Privacy Preserving In Federated Hybrid Smart Grid Cloud Environment Using Spark

Abstract

Association rule data mining provides to integrate large volume of dataset and extract information for making effective decisions. The extracting useful information is found large number of disjoint dataset with multiple cloud federations. Cloud federation is a collection of shared data and hosted from multiple cloud user environment. In this case the organizations are sharing data and adopt changing environments. The privacy preserving is an important factor and allows the organization to control the cloud environments while accessing the data. In the paper, we use association rule mining in federated cloud environment using spark model for preserving privacy. In this framework set of user can allows to access the private dataset and encrypt the dataset for providing privacy. In our proposed model we consider confidentiality, privacy and integrity major factors. We propose privacy preserving identity access mode with attribute based access control mechanism in cloud environment. The user can get access from set of policies granted by cloud environment. The spark model is used for testing the experiments and compares the results with existing methods.

Full text

European Journal of Molecular & Clinical Medicine
ISSN 2515-8260 Volume 07, Issue 07, 2020
195
Iot Enabled Privacy Preserving In Federated
Hybrid Smart Grid Cloud Environment Using
Spark
Dr. Shafali Jain1, Dr.A.Gnanasekar2, Dr S Hasan Hussain3, Dr.C.Bala Subramanian4,
S. Stewart Kirubakaran5
1Professor and Head, Department of Electrical and Electronics Engineering, Sagar Institute of
Research and Technology, Bhopal, Madhya Pradesh -462041.
2Associate Professor, Department of Computer Science and Engineering, R.M.D.Engineering
College, Kavaraipettai- 601206, Tamilnadu, India.
3Associate Professor, Department of Computer Science and Engineering, Syed Ammal
Engineering College, Ramanathapuram- 623502, Tamilnadu, India.
4Assistant Professor, Department of Computer Science and Engineering, Kalasalingam
Academy of Research and Education, Virudhunagar- 626126, Tamilnadu, India.
5Assistant Professor, Department of Computer Science and Engineering, Saveetha School of
Engineering, SIMATS, Thandalam-602105,Tamilnadu, India.
Abstract: Association rule data mining provides to integrate large volume of dataset and
extract information for making effective decisions. The extracting useful information is
found large number of disjoint dataset with multiple cloud federations. Cloud federation is
a collection of shared data and hosted from multiple cloud user environment. In this case
the organizations are sharing data and adopt changing environments. The privacy
preserving is an important factor and allows the organization to control the cloud
environments while accessing the data. In the paper, we use association rule mining in
federated cloud environment using spark model for preserving privacy. In this framework
set of user can allows to access the private dataset and encrypt the dataset for providing
privacy. In our proposed model we consider confidentiality, privacy and integrity major
factors. We propose privacy preserving identity access mode with attribute based access
control mechanism in cloud environment. The user can get access from set of policies
granted by cloud environment. The spark model is used for testing the experiments and
compares the results with existing methods.
Key terms: IoT, Cloud Services, Privacy Preserving, Smart Grid, Federations, Spark model
1. INTRODUCTION
The cloud federation is collaboration between various organizations that gives organizations
access to hosted data. Fixed cloud platforms for other federal members to take advantage of a
specific problem or specific business opportunity. However, the availability of cloud
federations is very important Adopting solutions that enable secure data sharing between
them Participating organizations. In particular, it is necessary Allowing identity and access
management systems Federal agencies to authenticate customers from other affiliates to
determine the companies and the permissions within them shared data.
The research team also suggested some identities Access management systems for federal
clouds. However, these systems have several limitations. First, they need trust agreements
between federations Companies to verify users' identities and map their access Manage

European Journal of Molecular & Clinical Medicine
ISSN 2515-8260 Volume 07, Issue 07, 2020
196
policies on shared data. These solutions are not ideal contributions for dynamic and Cloud
Federations Controls on the organizations. Above all, there must be partner organizations
Users' identities can be checked and access control enforced policies on shared data without
establishing trust agreements between them.
Second, it provides better access to specific solutions and Limit policies, partner with
potential competitors. Finally, different factors evaluate policies and implement the resulting
decisions. However, it is harmful users may damage access control policies or their
evaluation Engine to gain unauthorized access to shared data Federal. To this day, cyber-
attacks rob integrity no policy decisions and no implementation points Reported, but attacks
as witnesses are still possible Certification authority’s on user identification characteristics.
However, they do not protect the privacy of features, which convey strategic information that
companies are not ready for accepting services
One of the main goals of big data is to gather useful knowledge from large datasets in
practice Data Mining Techniques. An organization is effective mining for their data locally,
the demand for better data will increase mining results encourage companies to share their
data for mutual benefit. It is clear that data mining produces more accurate and useful results
when applied in different ways datasets collected from multiple organizations. This is the
process commonly known as Cooperative Data Mining (CDM). CDM can effectively address
many of today's applications in multiple domains such as social networks and healthcare
Finance, Construction, Cyber security, Biology, Physics. It requires the collection,
maintenance, integration and analysis of joint datasets. Similarly, it could be customer data
shared in insurance companies to improve their risk evaluation strategies. However, due to
increasing privacy concerns about data and increased government control Privacy and
companies may not share them confidential data with other organizations.
A solution Contribution to protecting data privacy is often argued Data mining is “data
mining that protects privacy”. However, Lack is a major drawback of existing solutions
scalability and limited applicability. Also, the majority they compromise security to increase
consumption Information. Companies want to source their data analytics work in our cloud
environment. Effectively address their financial and performance needs. Since the cloud is an
unreliable third party, data is often encrypted and then our source. In such a case, very limited
work Data analytics are performed in the cloud to protect privacy. In this paper has following
sections, section 2 discuss about various related works, section 3 gives models and procedure
for implementing privacy model, section 4 describes proposed model with simulation and
section 5 explain results and discussions.
Related Work
Cheng et al, one-to-one item mapping is suggested using a substitute technology to add
duplicate items to modified each transaction dataset. In particular, the data owner attaches
duplicate objects to dictionary, and then randomly maps each item to a subset values. As a
result, the server could not easily find the distribution in the correct item sets in the database.
Wong et al, this approach has two limitations. First, the possibility of adding counterfeit
goods to it every transaction is the same. I.e., fraudulent transactions appear in a large
database leading to similar frequencies cryptography analysis. Second, duplicate items were
added to transaction database is different for existing objects and so on some calculations are
true duplicate items can be eliminated and found. Myth et al, An approach based on K-
support is suggested anonymity to protect each sensitive species with k – 1 item in the
database with similar support value. Each the modified item will not be detected from at least
k – 1 Items.
Wait et al, the pseudo-taxonomy tree was used to limit their approach Incidents of counterfeit
goods. K-support suggests a model that extends the concept of anonymity to k-privacy, where

European Journal of Molecular & Clinical Medicine
ISSN 2515-8260 Volume 07, Issue 07, 2020
197
each modified element is at least k - 1 identifiable from other item sets to achieve K-privacy,
they first used the One to One mapping method to replace plain objects, and then Items are
divided into k-size groups for k-privacy. Sater et al, duplicate transactions have been added as
in the previous steps. Recently, Yi et al, Data is considered a problem the owner encrypts his
/ her database and stores it on a DB server.
The Consumer Association Rule Mining Task is sent n ≥ 2 are semi-standard servers, which
enforce the association rule enter the encrypted results by collection of cloud data into the
encrypted data User. They suggested three different solutions for data retrieval based on
canonity, K-support and K-privacy techniques. Chung et al, privacy is specific solutions
depend on the supplied Cryptocurrency system to obtain item privacy and transaction
privacy Database privacy.
However, their solutions are valuable the information is therefore considered insecure. The
proposed protocol addresses such security gaps. Federation Supervision includes runtime
monitoring and offline auditing, FRM and FSA components, respectively. The FRM is
formed by a set of distribution probes Controls all interactions between the federated during
runtime Clouds; If any error is found, security Increases the alert for the administration
console. FSA and more these interventions is audited offline using machine learning
Techniques. Federal management supports this work Management of cloud associations.
There is an IWM component Optimized workload strategies based on computing charge
Requests for implementation on members' cloud systems. The Provides administration
console instead of FAM component to control, regulate and monitor the condition of the
Confederacy. Through the RI component, all administrative data of important Confederation,
e.g. SLA and access control policies are stored Make the registry accessible to components
accordingly.
The Identity and Access Control Manager agrees our identity and access control management
system. The system offers the following features: Unknown identities. The system allows
access to shared data while protecting federal companies' customers Identification features
also for Access Control Manager. The Depending on the anonymity of user identities
(pseudo-) can be achieved Identity tokens without explicit submission of identity features.
Identity Manager gives users identification tokens Contains structured information. In
particular, each Username, identity attribute name, And a meaningful and secure Pederson
commitment the identity attribute value of the user signed by the Identity Manager.
Problem Statement and System Architecture
This section provides an overview of what we are referring to Identification and access
control system for cloud federation. As this system is intended to be part of spark Platform,
we will first introduce FaaS and its platform. FaaS is the equivalent of a service for security
enabled clouds Creation and maintenance of cloud associations. To this extent, It is supported
by the spark platform and its graphical description is reported in Figure 1 allows you to
federalize, verify cloud data and services.

European Journal of Molecular & Clinical Medicine
ISSN 2515-8260 Volume 07, Issue 07, 2020
198
Figure 1: Identify access control model for implementing privacy preserving – Smart Grid
Secure provisioning and access through trusted identity management and access control
functionality. In particular, the platform components reported in the image may be logical
Recognized as Federation Monitoring and Federation Management Identity and Access
Control Manager. Here we report that specific entities are a specific component Architecture.
Data Owner is readymade Federated Company Share data. Data Request User in the Cloud
Federation To access shared data. The Data Owner Program is a simple web-based
application. It gives the data owner an interface to share data with define access control
policies that control access such data. The Identity Provider Application (IDP) is cloud based
application that consists of three parts and runs on the Spark Enclave. It gets recognition
features from data request then generates identification tokens. Relay IDP is user-space
application Network connectivity from EnclIdP to CIdP Data Request Application. Interface to
data allows the app to verify Verification for EnclIdP is also sent to EnclIdP data the applicant's
identity characteristics and its request to issue identity tokens.
In our model, the following two entities are considered, (i). Set of users are represented U1,
U2,..Uq and (ii). Cloud environments are represented as C1 and C2. The transactional
database represented as Ti with respect to q users. Access Control Manager Application
(ACM) such as Identity provider application, ACM launched EnclIdP, An application that has
the following components: EnclockM is a program running in the Spark enclave Symmetry is
responsible for encrypting federal data using key encryption. It creates for each feature the
provision in the Access Control Policy is a special secret, this is called Conditional
Subscription Secret (CSS). The key is used to rebuild the federated decryption Information.
Enclock clearly communicates CMS to data Applicant running the collaborative association
mining with data EnclR of the request application through RelayACM. That too Keeps a
record of all deliveries of CSS. CACM is an excellent deal for storing AC access control
Policies and instructions defined by the data owner Encrypted data (simply, hash) is stored in
ACM.
Proposed Model with Implementation
Data Privacy - Specific protocol guarantees the contents of the user's database are never
disclosed for other users and for the Federated Cloud. Also, the final output is known only to
the participants Users. We demonstrate the security of our protocol is better than recent jobs
in visualised methods. User Accuracy - Our protocol always gives optimal results. It all
depends on the basic steps for the end user Standard compliment algorithm, but implemented

European Journal of Molecular & Clinical Medicine
ISSN 2515-8260 Volume 07, Issue 07, 2020
199
Encrypted data. Maximum cloud participation - one of the primary goals increasing the use of
our outsourcing cloud resources as much as possible. In our protocol, after users are
outsourced C1's their databases do not belong to any of them the remaining steps in the
specified protocol.
That is, all association rule mining activities. C1 and C2 work together so that it is max our
sourcing advantage to the cloud environment. It is worth it note that users may be offline
after our outsourcing they can also check their data and association rule mining Results when
they are needed. So the scalability most functions in a specific protocol are free and well
measured Supports large data applications.
Figure 2: Proposed association rule mining for implementing Smart Grid
The above figure 1 shows that proposed framework for entities among each cloud users. 1.
User u can outsource the data and set of association rule mining is applied in each C1 and C2.
2. The user group can holds all the information and set of rule mining is applied for each
transactional database values. 3. The encrypted database denoted as EnclIdP and Privacy
preserving collaborative association rule mining is applied by using below algorithm
The encryption function,
Enpk(a + b) = Enpk(a) * Enpk(b) mod N2
Whereas secure modelling is applied for input values a and b, r represented as random
number of each mask values
a * b = (a + ra) * (b + rb) – (a * rb)–(b * ra) – (ra * rb)
1. Secure data outsourcing Model
Our proposed model considered set of q user denoted as U1,..,Uq and apply association rule
mining with aggregated data. The Enpk is calculated as C1 and C2. The Enpk(T) is measured
as follows
Enpk(a,b) = 󰇛󰇜󰇛󰇜



2. Secure data from frequent dataset
The Encrypted dataset has support and confidence factor denoted as α and β. Set of user
U1,..,Uq and threshold values of each Enpk is obtained from outsource data factors.
Enpk(α) = 󰇛󰇜

 and Enpk(β) = 󰇛󰇜


So the confidence factor is obtained from Confidence(Enpk) = α * β / N
5. Experimental Results

European Journal of Molecular & Clinical Medicine
ISSN 2515-8260 Volume 07, Issue 07, 2020
200
We measure performance of our system using Java running on Ubuntu OS with 3.3 GHz Intel
i& processor, 8GB RAM. The transaction are randomly generated and following are
considered for Spark evaluation. Identity Token Issue, data request is provided by the
applicant Identity Properties to the identity provider, it provides an identity Token IT for each
identification feature. It has a nickname to uniquely identify the name in the system Identity
trait, Pederson commitment Identity attribute value. Each identification token is digitally
signed
Following the identity provider and data applicant the smart contract is stored in the chain by
CIdP. Signed Allows the identity provider to verify authenticity Identity token. When the
data is signed by the applicant After the identity is stored in the cloud dataset as required
tokens are publicly available. Without signature, another User identification tokens can be
introduced in the access controls Manage and obtain unauthorized access to federal data.
Identification token registration - once recognized Tokens, data can be requested from the
identity provider Request access to federated data in Access Control Manager. Then, to
retrieve the data decryption key, the data request must be submitted to the Access Controller
Manage identification tokens that match the feature Terms of the Access Control Policy
governing access Information. Then, Access Control Manager Requesting Conditional
Membership Privacy (CSS) data for Name each identifying attribute of a federated
organization Log in to the control system that matches the user identity Token name. User
uses CSS to obtain keys decrypt shared data that satisfies access Control mechanism. The
below table shows that input dataset values of different dataset values and token. The
threshold values are compared with existing methods.
Dataset
Token
Parameters
Iterations
Support and
Confidence
Threshold
Redshit
300
12
5
0.05
0.87
V3MD5
345
15
5
0.05
0.92
ShaF4
568
15
5
0.05
0.79
VmwT1
456
18
5
0.05
0.76
Classfish
347
21
5
0.05
0.89
Sunflash
127
24
5
0.05
0.91
Table 1: Dataset input applied in Spark model and threshold values are obtained from our
method
The same threshold values are compared with existing protocol and shown in table 2
Dataset
Support and
Confidence
Iterations
Convolution
Network
ToPo
Specification
VSphere
Method
Proposed
Model
Redshit
0.05
5
0.58
0.71
0.54
0.87
V3MD5
0.05
5
0.71
0.72
0.82
0.92
ShaF4
0.05
5
0.45
0.55
0.57
0.79
VmwT1
0.05
5
0.42
0.58
0.55
0.76
Classfish
0.05
5
0.45
0.64
0.71
0.89
Sunflash
0.05
5
0.41
0.71
0.81
0.91
Table 2: Comparison result of different protocol with our proposed model.
The spark generated results for computation time, threshold values and cloud user
representations. The following figure 3 shows the graphical results of each cloud user values
with respect to time and transactions.

European Journal of Molecular & Clinical Medicine
ISSN 2515-8260 Volume 07, Issue 07, 2020
201
Figure 3: Computation Time, Transaction and comparison results
2. CONCLUSION
Data mining gains strength in the cloud environment companies believe that cloud computing
can effectively meet their financial and performance needs processing large amounts of data.
We suggest in this article Privacy-friendly framework that allows a group of users to
cooperate with us on their data and association rule mining in a federated cloud environment.
We have shown the proposed protocol provides a strong security guarantee compared to the
current method. Also, we evaluated our protocol by demonstrating various protocols and
creating low costs among end users using spark. Our solution is secure standard semi-honest
model. As a future creation, we plan explore our protocol further and improve to get it
Security against malicious opponents. We do extend our research to other critical data
analytical task cloud encrypted data.
3. REFERENCES
[ 1] A. Almutairi, M. Sarfraz, S. Basalamah, W. Aref, and A. Ghafoor, “A distributed access
control architecture for cloud computing,” IEEE Softw., vol. 29, no. 2, pp. 36–44, Mar.
2012.
[ 2] F. Zhang, E. Cecchetti, K. Croman, A. Juels, and E. Shi, “Town crier: An authenticated
data feed for smart contracts,” Cryptology ePrint Archive, Report 2016/168, 2016,
http://eprint.iacr.org/2016/168.
[ 3] M. Singhal, S. Chandrasekhar, T. Ge, R. Sandhu, R. Krishnan, G. J. Ahn, and E.
Bertino, “Collaboration in multicloud computing environments: Framework and
security issues,” Computer, vol. 46, no. 2, pp. 76–84, 2013
[ 4] Bharath K. Samanthula, Salha Albehairi and Boxiang Dong, “A Privacy-Preserving
Framework for Collaborative Association Rule Mining in Cloud”, 2019 IEEE Cloud
Redshit V3MD5 ShaF4 VmwT1 Classfish Sunflash
0.58 0.71
0.45 0.42 0.45 0.41
0.71 0.72
0.55 0.58 0.64 0.71
0.54
0.82
0.57 0.55
0.71 0.81
0.87 0.92 0.79 0.76 0.89 0.91
Convolution Network ToPo Specification
VSphere Method Proposed Model

European Journal of Molecular & Clinical Medicine
ISSN 2515-8260 Volume 07, Issue 07, 2020
202
Summit, 978-1-7281-3101-6/19/$31.00 ©2019 IEEE DOI
10.1109/CloudSummit47114.2019.00025
[ 5] X. Yi, F.-Y. Rao, E. Bertino, and A. Bouguettaya, “Privacy-preserving association rule
mining in cloud computing,” in Proceedings of the 10th ACM symposium on
information, computer and communications security. ACM, 2015, pp. 439–450
[ 6] C.-H. Tai, P. S. Yu, and M.-S. Chen, “k-support anonymity based on pseudo taxonomy
for outsourcing of frequent itemset mining,” in Proceedings of the 16th ACM SIGKDD
international conference on discovery and data mining. ACM, 2010, pp. 473–482
[ 7] A. Machanavajjhala and J. P. Reiter, “Big privacy: Protecting confidentiality in big
data,” XRDS: Crossroads, The ACM Magazine for Students - Big Data, vol. 19, no. 1,
pp. 20–23, Sep. 2012.
[ 8] B. K. Samanthula, Y. Elmehdwi, and W. Jiang, “k-nearest neighbor classification over
semantically secure encrypted relational data,” IEEE Transactions on Knowledge and
Data Engineering (TKDE), vol. 27, no. 5, pp. 1261–1273, 2015.
[ 9] S Manikandan, K Raju, R Lavanya, R.G Gokila, "Web Enabled Data Warehouse
Answer With Application", Applied Science Reports, Progressive Science
Publications, E-ISSN: 2310-9440 / P-ISSN: 2311-0139, DOI:
10.15192/PSCP.ASR.2018.21.3.8487, Volume 21, Issue 3, pp. 84-87, 2018
[ 10] M. Hogan, F. Liu, A. Sokol, and J. Tong, ”NIST Cloud Computing Standards
Roadmap”, 2011, https://www.nist.gov/sites/default/files/ documents/itl/cloud/SP 500
293 volumeII.pdf.
[ 11] Shorouq Alansari, Federica Paci, Andrea Margheri, Vladimiro Sassone, “Privacy-
Preserving Access Control in Cloud Federations”, 2017 IEEE 10th International
Conference on Cloud Computing, 2159-6190/17 $31.00 © 2017 IEEE DOI
10.1109/CLOUD.2017.108