ISSN 2278-3091

Volume
Alpesh Vaghela et al., International Journal of 10,
Advanced No.5,
Trends in September - October
Computer Science 2021 10(5), September - October 2021, 2930 – 2935
and Engineering,
International Journal of Advanced Trends in Computer Science and Engineering
number of information patterns [1]. Figure 1 shows a different
ABSTRACT Available Online at http://www.warse.org/IJATCSE/static/pdf/file/ijatcse011052021.pdf
ABSTRACT method for a different phase of the big data life cycle.
https://doi.org/10.30534/ijatcse/2021/011052021

Development of Privacy-Preservation of Big Data with

support of Hyperledger Fabric and IPFS
Alpesh Vaghela1, Anilkumar Suthar 2
1
Research Scholar, Department of Computer Science, Gujarat Technological University, Gujarat, India and
Programmer, GMB Polytechnic, Rajula Gujarat, India,
[email protected]
2
Director, New L J Institute of Engineering and Technology, Gujarat Technological University, Gujarat, India,
[email protected]

Received Date : August 05, 2021 Accepted Date : September 14, 2021 Published Date : October 06, 2021

Volume, velocity, and variety are the 3V problems of big data

ABSTRACT processing. Big data analytics research focuses not just on
solving 3V problems, but also on balancing privacy and
Academics and industry researchers alike find security concerns. If researchers are unable to develop a
privacy-preservation of large data to be a very intriguing field solution for privacy and security, big data concepts and
of study. Data collection, storage, and processing are the three implementation will be rendered useless. Big data, for
steps of big data's life cycle. At different stages of the big data example, may save the health-care industry up to US$450
life cycle, different privacy and security solutions are used. billion if it is used in a methodical manner. Patients'
Many health-care stakeholders are working together to information has been separated into sensitive and
develop a new pattern for safeguarding people from an non-sensitive sections. When sensitive data is used in health
unknown disease while also promoting economic prosperity. care research, privacy issues arise. However, when big data is
The methods of big data processing and big data analytics will
employed in conjunction with privacy and security, a
be employed to discover new economic growth patterns.
structured programme will be used to forecast for the safety of
Because the current method of data anonymization leads to
county residents [2]. Big data is a type of data that includes
data breaches, researchers needed to develop a new way of
large data mining or knowledge discovery in databases both structured and unstructured information [3]. Data
(KDD), in which numerous parties share their data to identify security and data privacy are two distinct ideas. Data security
new patterns. This study introduces a novel way for data refers to the confidentiality, availability, and integrity of data,
mining privacy protection based on Blockchain and the as well as ensuring that data is not accessed by unauthorized
InterPlanetary File System (IPFS) (PPDM). The authors individuals.
propose leveraging Blockchain and IPFS to create the
ChainPPDM approach for preserving big data privacy. The
data saved on the blockchain is immutable, transparent, and
safe, and it allows for decentralized storage. IPFS is a
distributed file system that stores data in a decentralized
manner.

Key words: Big Data, Blockchain, Hyperledger Fabric, IPFS,

Privacy Preservation, Security

1. INTRODUCTION

Big Data is a favored research subject and hot topic for

industrial researchers and academicians around the world,
and it is fast developing in contrast to all other topics in the
research domain. Big data is one method for efficiently Figure 1: show the different methods for a different phase of the big
storing, processing, analyzing, and searching a massive data life cycle.

2930
Alpesh Vaghela et al., International Journal of Advanced Trends in Computer Science and Engineering, 10(5), September - October 2021, 2930 – 2935

Data security entails gathering and protecting data from role in the privacy-preservation of big data in the healthcare
unauthorized users, as well as destroying data that is no industry. Figure 4 depicts the roles of several actors in data
longer needed. The right use of data is characterized as data mining in the hospital setting.
privacy. Customers' personal data is used by businesses on the  Data Provider: In the privacy-preservation of big data in
condition that no personal information is disclosed and no the hospital area, all users of the hospital, including doctors,
data is used that is not useful to customers [4]. The concerns patients, nurses, pharmacies, and lab reports, who supply
of data abuse, such as privacy data transactions, are not some input in the form of data are known as data providers.
properly regulated due to a lack of legal limits and immature  Data Collector: A Data Collector is a hospital that provides
auditing tools, and the difficulty of maintaining privacy is a server and storage area for storing data provider data.
worsened. The keyword search feature needs to be performed Because all data providers must rely on the data collector for
over encrypted centralized data storage without disclosing their data privacy and security, the data collector is the most
any information about the search query or the retrieved important actor for privacy protection.
document in order to ensure data privacy. A  Data minor: A data minor is a technical expert who is
privacy-preserving keyword search is what it's called. Figure familiar with a variety of algorithms for discovering
2 shows the list of users who have their privacy protected by meaningful information.
the centralization data mining server. The requirements for  Decision maker: For economic growth, the decision-maker
preserving users' privacy from the centralization big data is obliged for publishing the optimal report from data
mining server are listed below. analysis and data mining.

Figure.2: shows the list of privacy-protecting of users from the

centralization data mining server. Figure 3: shows the blockchain method for all phases of the big
data life cycle.
 Data privacy- In the centralization big data mining server,
outsourced patient data is secured without revealing any
private or sensitive information to unknown third parties.
 Index Privacy- The security of the hospital's big data
infrastructure must be ensured so that other hospitals cannot
steal private data from centralization-based indexes.
 Keyword privacy- The centralization server of hospital
management does not steal any particulars data collection,
index, and encrypted keywords.
 Trapdoor unlikability- Developed a random trapdoors
function to safeguard data and fire privacy, as well as
ensuring that each trapdoor generated query is unique and Figure 4: show role of the different actors in hospital area data
that no one can connect the trapdoors [1]. mining
2. LITERATURE REVIEWS
We created novel healthcare data mining algorithms based on
blockchain, which is a decentralised storage system with The existing PPDM strategies have been thoroughly
immutable transaction properties, and IPFS, which is a examined and classed based on their methods, which include
decentralised storage system for huge data. The blockchain data manipulation procedures, which seem to be the current
method is depicted in figure 3 for all phases of the big data life technique's main downside. [5], [6].
cycle. Data Provider, Data Collector, Data Minor, and
Decision Maker are the four key actors involved in the To solve the problem of medical data sharing among network
privacy-preservation of big data mining, and they all play a hospitals, the authors propose the MeDShare system.
2931
Alpesh Vaghela et al., International Journal of Advanced Trends in Computer Science and Engineering, 10(5), September - October 2021, 2930 – 2935

MeDShare is a verifiable method for data auditing and control relevant knowledge from it during the data mining process.
of medical shared data in cloud repositories for big data At this time, all PPDM approaches are based on a centralized
databases, based on blockchain technology. Cloud service and distributed data-based system. Data suppliers should put
providers and other data guardians can achieve data their trust in a third party or a centralized system authorized
provenance and auditing while sharing medical data with person. We presented a decentralised blockchain-based
entities like research and medical institutions with minimum method for large data privacy preservation, in which data is
risk to data privacy by utilizing MeDShare. Users can access securely transmitted from data collector to data miner.
data using their pre-generated private key, and if the key is Because blockchain's data storage capacity is restricted, we
legitimate, the user can access data; otherwise, the transaction use HDFS for data mining. Private data is saved on-chain,
is declared invalid [7]. The authors [8] suggested framework while non-sensitive information or data required for data
provides uniform admittance standards for these records and mining is stored off-chain. ChainPPDM is a
safe record storage based on blockchain. Because blockchain permission-based decentralised system with excellent data
does not have enough storage capacity for huge data, the study mining feasibility and security.
utilizes an IPFS system to store off-chain storage. And the
role-based access also benefits the system as the medical 3. BLOCKCHAIN STRUCTURE
records are only available to the trusted and related
individuals. A Blockchain is made up of a database of transactions linked
by a cryptographic hash. To store enormous amounts of
The blockchain technology creates a distributed ecosystem medical data, however, a big data storage system with a
with decentralised and tamper-proof records, as well as a new blockchain is required. [12]. The blockchain's fundamental
way to secure and share electronic health record systems. The design is depicted in Figure 5, which contains an individual
authors suggest a new electronic healthcare data sharing block with a block header and a block body that is linked to
system based on blockchain that verifies data integrity. The the next to continue blocks. Each hospital has its own servers,
author created a smart contract that allows users to access, and each doctor associated with that hospital has his or her
alter, and remove healthcare records stored in the cloud. This own Blockchain network, which is connected to the others
solution employs a private key that is kept in the blockchain, [13]. The entire network linked to the system manager will be
and an authorized individual can use this key to access any in charge of system management.
cloud-based record using the smart contract [9].

The BlockDBM approach was presented by the authors for

decentralised trust management systems for various tasks
such as data gathering, data sharing, and storage for
processing through permission blockchain-based smart
contracts. The authors [10] used ways to improve blockchain
technology's trustworthiness. For data privacy and security,
blockchain technology eliminates the need for a third party.
BlockDMB is a secure storage solution that uses both Figure 5: The basic execution structure of blockchain.
on-chain and off-chain storage mechanisms to keep parties'
trust. Blockchain is very limited storage capacity as well as 4. PRIVACY PRESERVATION OF BIG DATA-
slow processing power. Blockchain as storage for traditional PERMISSION BC AND IPFS
big data systems is not possible. The authors [11] created a
blockchain-based personal management system. Using the The authors [14] proposed our novel schema and novel
on-chain and off-chain data storage methods, the author architecture, ChainPPDM, privacy preservation of big data,
devised unique data-based storage to solve the problem of data or privacy preservation of data mining. ChainPPDM is the
redundancy and insufficient storage capacity. The authors core four technology used in implementation parts i.e a)
found a solution to the problem of personal management Permission Blockchain - Hyperledger Fabric, b) HDFS, c)
system security in big data. The authors used a standard IPFS, d) SHA256 algorithm. Hyperledger Fabric can handle
database-managed approach to overcome the blockchain 3000 transactions per second. The ECDSA technique is used
challenge of data redundancy and storage. in the Hyperledger Fabric membership management, which is
based on a conventional X.509 certificate. Each member
This concept is expanded and improved, and a new receives a digital certificate through the PKI system. The
ChainPPDM for Big Data privacy preservation is proposed. Gossip protocol can only distribute data between nodes in the
Many anonymization and encryption approaches have been same MSP on the channel. MSP guarantees that access to the
developed to protect the privacy of big data while collecting Fabric platform is secure.

2932
Alpesh Vaghela et al., International Journal of Advanced Trends in Computer Science and Engineering, 10(5), September - October 2021, 2930 – 2935

ChainPPDM presents new on-chain and off-chain solutions inserted into the HDFS system of data minor. Figure 6 shows
for protecting big data privacy. We believe, however, that the the process of data set generation and stored in the blockchain
big data analysis method is faster than a database system. and centralized database.
Authors should keep bid data in mind while mining data or
looking for patterns. The most serious issue is data privacy.
For data analysis, ChainPPDM employed the HDFS big data
system and a blockchain system for user privacy. Healthcare
has information on customers who placed orders online. The
Hospital's patient database contains both confidential and no
sensitive information, such as the patient's address,
neighborhood, and city. ChainPPDM is kept private in a
Hyperledger Fabric collection, and non-sensitive data is kept
in HDFS. To send data from the data collector database to the
data minor warehouse, ChainPPDM employs IPFS.

4.1 Hash Value Generated and Data Transfer to Figure 6: the process of data set generation and storage in
Blockchain blockchain and centralized database

From the entire amount of data, the data collector has 5. ARCHITECTURE AND ALGORITHMS
constructed two independent sets of data. The patient's private
data is one set, while nonsensitive data from data mining is IPFS is utilized by ChainPPDM to send data from the data
the other. collector to the data minor quickly and securely. ChainPPDM
is implemented using Chaincode, MSP, statedb, IPFS, and
(1) HDFS. CouchDB is a state database that keeps track of
transaction results. Only core data is stored in the blockchain
is private data attributes of consumer network, while non-core data is stored in a central database.
The MSP is a pluggable interface that consists of a set of
(2) encryption methods and protocols for issuing and verifying
certificates and identities in the blockchain. The CA is used to
is nonsensitive data attribute of consumer and is whole data generate certificates and secret keys, as well as to initialize the
set. The data collector is created data set for data mining. MSP, and the order node serves as a network proxy for data
Before giving data set to transfer to Data Minor warehouse, distribution. Figures 7 and 8 depict ChainPPDM's
Data collector is decided sensitive data attribute for data architecture and algorithms, respectively.
privacy. To measure hash value, we use

(3)

where is the private data of the patient and we add some salt
to generate hash because every same input of value for a hash
function is generated the same value. It has a good chance of
guessing a private value from similar hashes. For data of any
length, the SHA256 algorithm can generate a 256-bit long
hash value [15]. A hash value is a numerical representation of
a piece of data that is unique and exceedingly compact.
Generated has become essential for both sensitive and
non-sensitive data. Now we create new data set using has
Figure 7: shows the architecture of ChainPPDM
value, . New data set has key and value pair. Hash
value becomes key for access to private as well as
6. EXPERIMENTAL RESULTS AND ANALYSIS
non-sensitive value.
We created a prototype system based on the Virtual Machine for
(4) ChainPPDM and the Hyperledger fabric to ensure that our data
separation and storage solution is genuinely effective for privacy
(5) preservation of big data using blockchain systems like data mining.
is inserted into blockchain private collection and is The system runs on an Ubuntu 20.04 (64-bit) virtual machine,
2933
Alpesh Vaghela et al., International Journal of Advanced Trends in Computer Science and Engineering, 10(5), September - October 2021, 2930 – 2935

Intel(R) Core (TM) i7-4790 CPU @ 3.60 GHz processor and 8 GB Figure 10 depicts Write Latency(second) plot using
RAM, and uses the HDFS simulation central database. We observe Hyperledger Caliper tool. In comparison to the BlockBDM
the changes in the storage capacity after separately sensitive and and PBMS, ChainPPDM function better because
non-sensitive data with the hash value of non-sensitive data. Table 1
ChainPPDM write private data in private state of Hyperledger
shows compression between other method for privacy-preservation
of big data with ChainPPDM. Hyperledger Caliper tool has been
Fabric.
used for evaluation of ChainPPDM.

Figure 9: Read Latency(second) plot using Hyperledger Caliper tool

When data is minimal, uploading takes longer than cloud

uploading, but as data grows, cloud speed suffers.
ChainPPDM has a larger storage capacity, but it takes less
time to do a data mining process while maintaining
anonymity.
Figure 8 shows the Algorithms of ChainPPDM

Table 1: shows compression between other method for

privacy-preservation of big data with ChainPPDM.

BlockBD
PBMS [9] ChainPPDM
M [8]
Public
Ethereum NA NA
Blockchian
Permissioned hyperledge hyperledge hyperledger
Blockchain r Fabric 1.4 r Fabric 1.4 Fabric 2.0
Data Type multimedia text text Figure 10: Write Latency(second) plot using Hyperledger Caliper
Database tool.
before MySql MySql MySql
system 7. CONCLUSION
Nodes
4 4 4 ChainPPDM has solved the challenge of massive data privacy
Number
On-Chain preservation. ChainPPDM presents a blockchain-based
Blockchain Blockchain Blockchain architecture for large data privacy preservation, as well as a
Storage
Off-chain prototype system to demonstrate feasibility and provide a
IPFS MySql IPFS, HDFS
Storage solution for big data mining technologies. In addition, data
was separated into two sections, which increased data privacy
The average size of a single data provider record in single and solved the problem of greater data computation in
files is 182 bytes, and the data provider separates this data into blockchain vs traditional data mining systems. In this article,
two sections, with private data taking up 142 bytes and we will just utilise text big data to create the system; however,
nonsensitive data taking up 40 kb. The storage load of we will assess it and try to extend it to additional fields in the
non-sensitive hash data grew with byte kb of each record on future.
blockchain as well as on centralized data minimal storage.
Figure 9 depicts the Read Latency(second) plot using ACKNOWLEDGEMENT
Hyperledger Caliper tool. BlochBDM and PBMS read latency
low compare to ChainPPDM because ChainPPDM is used The authors are thankful to the GMB staff members, Dr.
private state of Hyperledger Fabric. With the secrecy of data, Nalin Jani, KSV and Dr. Seema Mahajan, IU for their best
ChainPPDM increases file size on hard disc. cooperation, support and guidance.
2934
Alpesh Vaghela et al., International Journal of Advanced Trends in Computer Science and Engineering, 10(5), September - October 2021, 2930 – 2935

REFERENCES 14. A. Vaghela and A. Suthar, A review of Bigdata analysis

1. P. Sreekumari, Privacy-Preserving Keyword Search using smart contract, Juni Khyat Journal, vol.X,
Schemes Over Encrypted Cloud Data: An Extensive issue.XII, pp 109-115, 2020.
Analysis, in 2018 IEEE 4th International Conference on 15. S. H. Begum and F. Nausheen, A comparative analysis
Big Data Security on Cloud (BigDataSecurity), IEEE of differential privacy vs other privacy mechanisms
International Conference on High Performance and for Big Data, in 2018 2nd International Conference on
Smart Computing, (HPSC) and IEEE International Inventive Systems and Control (ICISC), 2018.
Conference on Intelligent Data and Security (IDS), 2018.
2. R. Lu, H. Zhu, X. Liu, J. K. Liu and J. Shao, toward
efficient and privacy-preserving computing in big
data era, IEEE Network, vol. 28, pp. 46-50, 2014.
3. D. Zhang, “Big Data Security and Privacy Protection,” in
Proceedings of the 8th International Conference on
Management and Computer Science (ICMCS 2018),
2018/10.
4. A. Mehmood, I. Natgunanathan, Y. Xiang, G. Hua and
S. Guo, Protection of Big Data Privacy, IEEE Access,
vol. 4, pp. 1821-1834, 2016.
5. Q. Xia, E. B. Sifah, K. O. Asamoah, J. Gao, X. Du and
M. Guizani, MeDShare: Trust-Less Medical Data
Sharing Among Cloud Service Providers via
Blockchain, IEEE Access, vol. 5, pp. 14757-14767,
2017.
6. A. Shahnaz, U. Qamar and A. Khalid, Using
Blockchain for Electronic Health Records, IEEE
Access, vol. 7, pp. 147782-147795, 2019.
7. S. Wang, D. Zhang and Y. Zhang, Blockchain-Based
Personal Health Records Sharing Scheme with Data
Integrity Verifiable, IEEE Access, vol. 7, pp.
102887-102901, 2019.
8. M. Zhaofeng, W. Lingyun, W. Xiaochang, W. Zhen and
Z. Weizhe, Blockchain-Enabled Decentralized Trust
Management and Secure Usage Control of IoT Big
Data, IEEE Internet of Things Journal, vol. 7, pp.
4000-4015, 2020.
9. J. Chen, Z. Lv and H. Song, Design of personnel big
data management system based on blockchain, Future
Generation Computer Systems, vol. 101, pp. 1122-1129,
2019.
10. R. Kumar, N. Marchang and R. Tripathi, Distributed
Off-Chain Storage of Patient Diagnostic Reports in
Healthcare System Using IPFS and Blockchain, in
2020 International Conference on COMmunication
Systems NETworkS (COMSNETS), 2020.
11. C. Rahalkar and D. Gujar, Content Addressed P2P File
System for the Web with Blockchain-Based
Meta-Data Integrity, in 2019 International Conference
on Advances in Computing, Communication and
Control (ICAC3), 2019.
12. The Hyperledger Project. [Online]. Available:
[Online]. Available: http://www.Hyperledger .org..
13. D. Rachmawati, J. T. Tarigan and A. B. C. Ginting, A
comparative study of Message Digest 5(MD5) and
SHA256 algorithm, Journal of Physics: Conference
Series, vol. 978, p. 012116, 2018.

2935