Content uploaded by Qin Wang
Author content
All content in this area was uploaded by Qin Wang on Feb 29, 2020
Content may be subject to copyright.
ArtChain: Blockchain-enabled Platform for Art Marketplace
Ziyuan Wang, Lin Yang, Qin Wang, Donghai Liu, Zhiyu Xu, Shigang Liu
Blockchain Innovation Centre
Swinburne University of Technology
Abstract—Blockchain is an emerging technology that has
the potential to revolutionize the global industry and create a
trusted relationship in a multi-party business network. There
are a number of practical use cases where blockchain has been
applied. One specific area is the Art industry, where it is a
natural fit in the way that art forensics and transactions are
conducted, tracked and recorded. This motivates us to develop
the ArtChain platform to assist the Art Industry. In this paper,
we present ArtChain, which is an integrated trading system
based on blockchain. It includes the front end, the back end,
the services, the smart contract, the chain connection and the
deployment scripts from the bottom to the top. To the best of
our knowledge, this is the first deployed blockchain-enabled
art trading platform in Australia. It provides a transparent
yet privacy-preserving, and tamper-proof transaction history
for registration, provenance, and traceability of art assets.
Our objective analysis and evaluation show that the ArtChain
platform is applicable and practical. For the interest of other
researchers, our system implementation related resources are
open-sourced on Github1.
I. INTRODUCTION
Blockchain, also known as distributed ledger technol-
ogy (DLT) [1], is designed to support verification-driven
transaction services within a generally un-trusted ecosystem.
The design of blockchain technology ensures that no one
business entity can modify, delete, or even append any
record to the ledger without consensus from other network
participants, ensuring the immutability of data stored on the
ledger. Blockchain is now being used in several industry
applications such as blockchain-enabled traceability and
provenance for food safety [2] documentation and cross-
organization workflow management in trading and logistics
[3].
With $200 billion of annual trading, the art market is one
of the largest unregulated markets in the world, accounting
for one-third of the amount of crime just behind drugs and
guns [4]. Tens of millions of dollars are transferred with little
or no documentation and transparency. Current challenges
and issues in the art market are: (1) lack of transparency on
prices and ownership history (provenance) and inadequate
control of transaction data due to the information asymme-
try; (2) the authenticity and appraisal of high-value works of
art is difficult; (3) lack of the value of artworks at the primary
art market and transparency trading at the secondary auction
market (both online and offline); (4) lack of recognition,
1https://github.com/ArtChainGlobal
public attention and care for a large number of artists; (5)
it is difficult for the artists to get royalty payment from the
secondary market.
Blockchain technology possesses a natural fit to improve
the transparency, keep records and reduce illicit activities
in the art market, due to its inherent properties [5] [6]. In
this paper, we present our project work, called ArtChain,a
blockchain-based art trading system, which has been piloted
and operated as a working product in practice. It is expected
to provide a complete solution towards these challenges by
creating a new ecosystem for the art keeping, trading and
transferring. ArtChain fundamentally builds up the underly-
ing architecture of blockchain to support a commercial-level
trading platform centered around art assets. The core value
proposition of the platform lies in:
•Privacy Protection Shared ledger along with permis-
sioned control ensures the transparency of each trans-
action which guarantees the privacy protection in art
trading and provenance.
•Traceability Real-time tracking of individual artworks
combined with the blockchain ledger assists in the fight
against counterfeit artworks.
•Irreversibility The on-chain registration of collectors
offline assets provides an immutable digital record of
the artwork, which guarantees the true ownership, the
provenance and the value of the artwork.
•Transparency Publicly displaying artworks to a wider
range of professional investors, leveraging the openness
of art ecosystem.
II. BLOCKCHAIN SOLUTION TOWARDS ART TRADING
In this section, we start with the rationale behind the use
of blockchain for an art marketplace, and then discuss the
benefits of this blockchain-enabled platform.
A. Rationale behind Using Blockchain
The major entities or participants in our solution are
described in the following.
•Artist Established artists along with new generation
artists all have equal opportunities for professional
evaluation and to publish their artworks. Published
items will be available for trade.
•Art gallery After artwork is registered on the blockchain
system, it can be tracked and located in real-time giving
an additional level of security to galleries.
447
2019 IEEE International Conference on Blockchain (Blockchain)
978-1-7281-4693-5/19/$31.00 ©2019 IEEE
DOI 10.1109/Blockchain.2019.00068
'%-"
()')%$
%$
)"(
('#$'%$)$
'$ ')$
%"")%$
')
,"")
%$)$)
')+%'!$#$)
()'
"(
+$'(&'-)%$
+$'(&$$
%&&$$
"")$#$)
"$
'$('
'$()%$
)'+
%"
%,"%"
+'$
%$)'%""'
(
#')%$)')
)'*#
Figure 1. System architecture
•Auction house Data stored on blockchain can be syn-
chronized with auction houses, opening up new chan-
nels for a greater global audience participation. More
traffic equals more opportunities for all involved.
•Collector Online synchronization of collectors offline
art assets. This proves ownership and sets the prove-
nance of the pieces for future generations ensuring the
value is preserved. Access to a global database with
extensive filtering capabilities.
B. Benefits of a Blockchain-enabled Art Marketplace
Firstly, the artwork authenticity and traceable data can be
simply achieved. Provenance is crucial when it comes to
collecting art. Not having a record of the ownership history
for a masterpiece often raises suspicion that it could be
stolen or fake, hence a distributed ledger can be used to trace
the transfer of ownership over a period of time, and serve as
a decentralized database securing provenance data and other
important information related to artworks. This allows for
quick and indisputable ownership transfer in trading.
Secondly, royalty payment from the secondary market for
artists can be achieved. A 5% royalty will be payable to
visual artists on certain commercial sales of their work. This
entitlement is created by the Resale Royalty Right for Visual
Artists Act 2009.However, due to the difficulty in tracking
the resales in the current art market, often artists do not
necessarily get the royalty [7].
Thirdly, Blockchain audit trail helps in detecting tax
evasion and money laundering. Add-on analytics or AI
services can predict the current value of an artwork based on
shared transparent data. This helps primary market valuation,
which is more difficult and more speculative than secondary
market due to a lack of market history.
Furthermore, our solution is designed to become an open,
expandable infrastructure orientated towards the art industry.
This means that participants will have the opportunity to
develop an extensive range of art-related applications for
specific scenarios based on the foundation of ArtChain.
III. SYSTEM OVERVIEW
In this section, we first present the foundational principles
the architecture is based on, the high-level architecture and
its main components. Then, we present basic data model
design and the trading process of the platform. In addition,
we discuss the trust and security issues.
A. High-level Architecture Design
We first evaluate several blockchain platforms to inform
our decision on which platform to apply. Based on the
business requirements and technical assessment we decide to
use the Ethereum private blockchain and Proof of Authority
(PoA) [8] as the consensus algorithm. Initially, we consid-
ered to use Hyperledger Fabric to implement our system due
to its capability, popularity and maturity. However, it lacks
support in native token, which is a key business requirement
in our design as the art trading platform hopes to integrate
the payment process and the ownership transfer process. We
design and implement a utility token called ACGT to achieve
the high performance requirement. Refer to Section IV-A
Tokenization for more details.
Here, we adopt microservices architecture for the fol-
lowing benefits: (1) Allows quick parallel development of
various components in the application landscape; (2) Re-
duces discussion time between various groups developing
various components; (3) When done properly, provides
clean reusable interfaces; (4) When done properly, reduces
handshaking in interfaces; (5) Reduces the risk and time of
integration/chain testing. The architecture design is shown
in Figure 1.
There are three layers in the system: the user front end,
the trading back end, and the ArtChain blockchain layer.
•User Front End: includes the following functions:
managing Profile for user registration, login and user
details; displaying art Collection; shopping Cart; user
Wallet; and CMS (Content Management System) to
create and manage web content.
•Trading Back End: consists of Artwork Management,
Shopping/biding, and Wallet Management. Artwork
Management includes artwork registration, ownership
verification and ownership transfer. Artists or collectors
conduct the registration of their artworks through the
assessment system of professional institutions within
ArtChain. Their works of art will then be eligible for
trading and participating in the ecosystem.
•ArtChain: including the following components: (1)
Royalty model: responsible for artists royalty payment
in the resale of their artworks. (2) POI model: manage
Proof of Interaction (POI) agreements, which are used
as incentives to grow the ecosystem of applications.
448
Figure 2. Data model: User Class
More details are described in Section IV-A. (3)Re-
warding Controller: based on POI model to manage
the rewarding to participants. The details are business
confidential information, which is out of the scope of
this paper.
B. Design of Data Model
There are three major groups of data objects stored in the
distributed ledger as illustrated as follow:
•User: contains all information related to a user’s profile,
login, wallet, and auction events attended. An artist is
also a user, with additional information and verification.
The detail is shown in Figure 2
•Artwork: consists of details, tag, history of ownership
transfer, and order details. These class represents the
workflow related to the masterpiece. The detail is
shown in Figure 3.
Figure 3. Data model: Artwork Class
Figure 4. Data model: Trading Class
•Trading: combines an order with the artwork and the
buyer’s basic information and the shipping address. The
detail is shown in Figure 4.
The trading process is shown in Figure 5. When the user’s
trading request is received, Profile Services and Trading
Services are triggered to retrieve the customer info and
trade info. After checking the trading conditions, Payment
Services are responsible for handling payment. Then Reward
Services are called to request and receive the reward infor-
mation. In the end, Shipping Services handle the shipping
information.
C. Trust Establishment
ArtChain co-operates with specialized or high-profile part-
ners in the primary or secondary markets of the art industry,
including museums, art galleries, and auction houses, to
establish the original ledger nodes and provide core func-
tions such as validating, ordering and generating blocks of
transactions. These ledger nodes and other agent or routing
nodes work together to protect the blockchain network.
We use Proof-of-Authority (PoA) as the trust model of
ArtChain network. PoA is well-suited to regulated industries
where entities are responsible for maintaining the network
(known as authorities), rather than remain anonymous as in
mining-based chains.
For our practical purpose, well-known museums and art
galleries are acting as authorities in ArtChain network to
conduct authenticity and price assessment for an artwork.
They are called SuperNode. Supernodes perform validating,
block generation and publishing.
449
%#"&$%
#%#
#%( ### $$ ( #'$
"&$%'
#
!'
# +
#'
$%#%
"&$%'
$!!
"&$%'
%#
"&$%'
&$%
* #
#
#'
!!
#'
*%
#'
(#
#'
&&#
)!#$$
Figure 5. Trading Process
Any node attempting to engage in malicious conducts
or falling under attack will be immediately detected by
other nodes in the network once it shows unusual behaviour
(e.g., sending illegal transactions, traffic attacks, and data
tampering). The network will immediately isolate the partic-
ular node and send out warnings. ArtChain deploys ledger
nodes throughout the primary and secondary art markets,
including internet companies, cloud service providers and
a large number of collectors of works of art and artists,
which, from a probability point of view, can eliminate the
possibilities where the majority of nodes fall under attack
or collude to engage in malicious conducts.
Initially, we set up 100 nodes to provide sufficient re-
dundancy and fight against 51% attack. Currently, they are
deployed at AWS and Ali cloud and not activate all at the
same time. We monitor the nodes behaviour and dynamically
replace the crashed nodes. We plan to extend the deployment
to be more decentralized on other clouds. In this regard,
an important issue to consider is the trade-off between
decentralization and performance.
IV. SCALABLE BLOCKCHAIN IMPLEMENTATION
In this section, we present the implementation of ArtChain
network. We first introduce tokenization and how it works
in our system. Then we describe the design and imple-
mentation of the upgradable smart contract for the purpose
of improving function and fixing bugs. Finally, we discuss
how to preserve privacy and confidentiality as required by
regulations and business needs.
A. Tokenizaiton
Tokenization refers to converting an asset into a digital
token on the blockchain system, so that ownership of the
asset can be transferred via smart contracts. Smart con-
tracts have functions for automatic transactions, formulas
for calculating asset prices and other specific features [9],
[10]. Tokenization is not simply the creation of a token.
Instead, it is about the design of the whole system, including
understanding the various rights and issues.
There are two types of token in ArtChain: the security
token ACG2and the utility token ACGT.
ACG token comes with the essential technical features
of digital currencies, including a steady issue curve, free
trading, immunity to double-spending attacks, and traceable
transaction history. These features are secured through the
ledger architecture and smart contracts. We develop rele-
vant E-wallets for corporate or institutional users, which
incorporate all essential functions for interactions with the
applications on ArtChain.
ACG token provides incentives for maintaining the
ArtChain network and the ecosystem of ArtChain applica-
tions.
•Network Maintenance: the consistency of ArtChain net-
work is jointly assured by ledger nodes. Ledger nodes
will have the opportunity to be awarded with ACG
as block rewards and transaction fees, to encourage
them to contribute to the security and stability of the
ArtChain network.
•Ecosystem of Applications: ArtChain will award users
with newly added ACG in positive correlation within
a certain cycle based on a number of indicators such
as their frequency of interaction with the ArtChain
ecosystem, levels of contribution, influence and the
number of ACG coins they hold. All indicators of
ecosystem incentives are quantifiable and verifiable,
which are collected and calculated by ledger nodes.
Incentives will be allocated under Proof of Interaction
(POI) agreements.
The ratio of the incentives for ArtChain network main-
tenance and the incentives for the ecosystem of ArtChain
applications will be dynamically adjusted by using a negative
feedback mechanism to maintain the balance and stability of
the ArtChain network and the ecosystem of applications.
The specific indicators and algorithms will be published
before any main relevant applications go online, and will be
implemented and operated through open rules of contracts.
Relevant institutional users of the ecosystem (art galleries,
museums, auction houses and artists) will be consulted.
The utility token ACGT is only used internally to facilitate
payment in art trading. It is a kind of stable coins, which are
designed to have a stable price or value over a period of time,
therefore, less volatile. These coins aim to mimic the relative
price stability of fiat currencies on one hand, but keep the
core values of cryptocurrencies such as decentralization and
security, on the other hand. Each ACGT token is collateral-
ized by an equal amount of fiat currency (1 AUD) held by
2https://etherscan.io/token/0x984c134a8809571993fd1573fb99f06dc61e216f
450
Figure 6. Zepplin proxy architecture pattern
a central custodian. Holders are guaranteed to redeem their
token at any point for the stable value denominated in fiat.
B. Upgradable Smart Contracts
Smart contract, once deployed into the blockchain, is
immutable literally. In consideration of bug fix and function
improvement, lots of work has been done to propose an
upgradeable design pattern of the smart contract [11].
The typical methods include:
•Separate logic and data
•Partially upgradable smart contracts system
•Separate logic and data in key-value pairs.
•Eternal storage with proxy contract
Among these methods, the proxy mechanism is most
flexible and guarantees a 100% upgradable mechanism i.e.,
the logic could be completely modified while remaining the
existing data state. In our system, we refer to Zeppelin’s
proxy patterns [12] and implement the so-called Unstruc-
tured Storage Pattern. The contract structure is shown in
Fig 6.
By using this pattern, the system achieves following
features:
•General user is unaware about upgrade of the contract.
•Implementation of logic contract is 100% upgradable.
•Data is stored in proxy contract. New data fields
could be added by the upgraded logic contract, without
touching existing data structure.
This design has also been chosen for the ZeppelinOS
smart contract system, and gone through a full security audit.
C. Initializing Issue
In our implementation, we also address the initializing
issue. This is a long-standing problem of upgradability solu-
tions for Ethereum. Our aim is to create an upgradable logic
contract, and we practically deploy a proxy contract which
delegates to a pre-existing deployment of logic contract on
the blockchain. Therefore the proxy contract has no chance
to establish the initializing steps in the constructor of logic
contract, and thus we need to do something special in order
to correctly initialize the proxy contract.
Our workaround for it is to use an initializer function
instead of the constructor, and make sure it is only executed
once for a necessary initializing process. Other proposals
could be found as Initializer Contracts [13] [14].
D. Safety Control
Transaction security: ArtChain network assures the se-
curity of users’ accounts and funds by using blockchain
consensus, digital signatures and end-users encrypted wal-
lets. The artwork trading platform provides security services
that are likened to those offered by financial institutions. It
integrates data, applications and transactions in blockchain
clouds through the efficient integration of data storage, net-
work and other resources, so as to create a secure transaction
environment.
Financial Management: ArtChain maintains high stan-
dards of integrity and ethical business conduct and is in
compliance with relevant laws and regulations, as well as
self-regulatory principles of the industry. We also implement
a component to conduct the regulatory duty of Know-You-
Customer (KYC).
E. Privacy and Confidentiality
ArtChain makes public all ledger nodes and their state
in the network in real time. The transaction history (block
content) and state information in ArtChain are publicly vis-
ible. However, in case of any privacy requirements for some
transactions, such privacy information will be processed.
In the data model design, we carefully decide what data
to be stored on-chain and what off-chain. The design has
been evolved along business needs and regulatory needs.
Currently, the on-chain data store contains information on
artist, the hash of ownership, price, and history. The hash
of ownership protects the privacy for owners who do not
want to be known to the public, as well as in compliance
with privacy regulations, such as the General Data Protection
Regulation (GDPR)3.
V. P ERFORMANCE EVALUATION
We conduct an extensive performance testing of the
system. We identify that the performance bottleneck of the
system is the low-level I/O efficiency of the Ethereum client,
i.e. Geth4in our system.
ArtChain private chain, based on POA consensus and
5-second block interval and deployed on 6 cloud nodes
(8x2.5GHz CPUs, 32G memory), supports up to 1500 TPS,
i.e., 1500 raw transactions on the chain, far more superior to
Ethereum mainnet (about 15 TPS nowadays). Integrated user
actions, like post new artwork or top up tokens, are usually
comprised of a series of transactions/queries on the chain.
ArtChain on average processes about 40-70 user actions per
second.
3https://eugdpr.org/
4https://geth.ethereum.org/
451
A. Environment Setup
The private chain is composed of 3 Ali cloud servers (8x
2.5GHz CPUs, 32GB memory, 64GB hard disc), and 3 AWS
cloud servers (8x 2.5GHz CPUs, 32GB memory, 8GB hard
disc). Geth version 1.8.17, startup parameter is tuned as:
•--targetgaslimit 4294967295: increase the gas limit to
0xFFFFFFFF to seal as many transactions as in one
block. Note this need to coordinate with the gasLimit
in the genesis.json file when creating the chain.
•--txpool.lifetime 24h --txpool.accountslots 65536 --
txpool.globalslots 65536 --txpool.accountqueue 64 --
txpool.globalqueue 65536: increase txpool so that it
stores as many transactions both account specifically
and globally as we submitted.
This paper employ web3.js5to communicate with the
chain, and to monitor its performance, Wireshark6is applied
to capture packet for analysis.
B. Throughput Analysis
Basically, blockchain throughput is limited by: a) How
long it needs to generate a block, and b) How many
transactions can be sealed in a block. And theoretical
throughput = (number of transactions in a block)/(block
interval). But in a large-scale network, the throughput is
also restricted by the broadcast speed. An explicit example is
Ethereum mainnet, with network congestion, its throughput
dramatically degrades as nodes need more time to keep
synchronized. This is why Ethereum is considered to have
issues on network scalability.
As for our private chain, we tried following steps to
tune up the system performance: (1) Speed up the block
generation by changing the block interval when generating
the genesis.json. To summary, the chain with 1-second
interval shows the best performance, but 5-second is also
acceptable. (2) Improve the gas limit of the chain. It does not
shows significant improvement on the performance, because
the gas limit is not the bottleneck of the system.
As our chain is only maintained by 6 cloud servers, we
can ignore the effect of network scalability mentioned above.
As long as the transactions are sealed, the nodes always
have adequate time to keep sync. On the contrary, it is the
node’s hardware configuration that determines the system
performance. We observe frequently crash of Geth client
on the node with only 8GB memory originally. Using the
node with 32GB memory, the performance is significantly
improved, but the crash still occurs in certain scenarios.
Geth is thought as a memory monster whose design
follows a “I use up what I have” idea, and will use up
all available memory on the server. By default, our node
servers disable the swap and will kill Geth process if it tries
to use up the memory. Unfortunately, this always occurs.
5https://web3js.readthedocs.io/en/1.0/
6https://www.wireshark.org/
We observed it used up 8Gb memory when trying to create
70 new accounts. A suggestion is to enable the swap on the
node, in terms of the sacrifice of the performance. Note the
AWS cloud server has only 8GB disc space, and so the swap
space is restricted on AWS servers.
C. Test Results
1) Raw transaction test: The chain is configured with 5-
second block interval and we get that (1) Transaction carries
data of 50 bytes, it is a typical value for general transactions.
(2) Establishes 2000 transactions in about 6-8 seconds per
node. (3) Establishes 20000 general transaction queries in
2-5 seconds per node.
2) API based test: APIs such as check user(),
check transaction() and check artwork() only query
information from the chain and do not include any
practical transactions. So they show as high throughput
as general queries. It only depends on the processor
and network performance. APIs such as buy tokens(),
post new artworks() and f reeze tokens() combine a se-
ries of queries and transactions, and these operations usually
depend on the result of the precedent, so those APIs have
bad parallel performance. For example, post new artwork
includes 16 low-level operations:
•2x eth.sendTransaction
•1x personal.unlockAccount
•1x eth.estimatesGas
•2x eth.gasPrice
•6x eth.getTransactionReceipt
•2x eth.subscribe
•2x eth.unsubscribe
Some test results are listed below:
•Establish 116 times API post new artworks() within
about 18-20 seconds per node.
•Establish 58 times API buy tokens() within about 5-6
seconds per node.
API add new user() contains a low-level operation of
personal.newAccount, which uses significant memory and
CPU cycles. A typical result is listed below:
•Establish 64 times API add new user() within about
20 seconds per node.
3) Test on different block interval: We tested on different
block intervals of 1 second, 2 seconds, 5 seconds and 15
seconds. The comparison of API based throughput with
different block intervals is summarized in Figure 7.
As all the APIs are called at the same time during the
test, we observe actually all transactions are sealed in one
block. Our chain is fully capable of guarantee that. So be-
sides the block interval difference, those calling procedures
need almost the same processing time on the network and
processes. That is why the different block interval practically
results in different throughput.
452
%ORFNLQWHUYDOVVHFRQG
1XPEHURIRSHUDWLRQVQRGHVHFRQG
SRVWBQHZBDUWZRUN
EX\BWRNHQV
Figure 7. Test under different block interval
4) Test on node crash: Geth client crashes under cer-
tain scenarios. What we notice is that transactions like
personal.newAccount(included in API add new user())
make Geth consume lots of memory. Got in tests:
•On the node configured with 32GB memory, Geth could
support up to 70 concurrent add new user() calls.
•Geth crash when submitting more than 70
add new user() calls. It is killed by OS after
using up all 8GB memory.
We then try to enable 32GB swap on the server,
and find that Geth succeeds processing 96 concurrent
add new user() calls. During the process, it used up 32GB
physical memory, and then 9.1GB swap memory. As a result,
it uses as long as 914 seconds to establish all the calls.
As a comparison, it needs only about 20 seconds for 64
calls without using swap memory. Performance degradation
is obvious.
D. Bottleneck of the System
Based on our performance test, we find out that: (1) The
performance bottleneck of our system is at the Geth IO
execution. (2) The way to improve the system performance
is to improve node hardware configuration.
According to [15], Ethereum uses LevelDB as the
database to store key/value. The key to accessing database
is irregular on account of the discreteness of hash. The
LevelDB has an excellent performance in reading/writing
continuously, while bad for the random key. Therefore, the
time tfor accessing LevelDB would be longer as the amount
of data storage increases. In fact, the test results show that
if nis large enough, the value of twill increase and the
eciency will degrade largely for some data which not hit
LevelDB cache at times.
Geth consumes huge memory on certain transactions,
e.g., personal.newAccount, and will crash when receiv-
ing multiple concurrent memory-consuming transactions. A
suggestion is to enable swap memory on the node to improve
system stability, at the sacrifice of the performance (See
performance degradation when memory is swapped). We
suggest 4GB of swap space on the nodes, based on the
performance test result. This improves the system stability
and does not degrade the performance significantly.
VI. RELATED WORK
In this section, we mainly review the work that closely
related to this work, for more work about blockchain and
the related applications, please refer to [1], [16].
Art as Digital Assets: Arts can be regarded as the
digital asset to be stored on the blockchain platform. The
blockchain inherently holds the property of authenticity,
traceability, and irreversibility which can perfectly pro-
tect the digital assets for each masterpiece. Usually, the
blockchain-based solution marks each masterpiece with an
ID, may denote as token in smart contract. Similarly, many
protocols are designed to trade the nonfinancial assets in
form of tokens on the blockchain platform
Blockchain Solution: Since digital assets need proper-
ties both on authenticity and security, blockchain becomes
the primary selection for the requirements. There are three
options, including private blockchain, consortium blockchain
and public blockchain. Due to the high security of the
assets, the most suitable solution is the consortium methods,
which relatively has a better trade-off between performance
and security. The asset-based property is deployed on the
application-layer of the blockchain, regulated by the rules
defined in smart contract. There are plenty of applications
successfully executed on top of blockchain [17] and sub-
sequently the infrastructure [18] becomes more complete
along with the development. Our solution provides a trading
infrastructure for art, and it provides an paradigm for other
high value commodities.
Privacy Protection: For the precious digital art assets,
it is fundamental to protect the privacy of assets. There
are two kinds of privacy in research, including identity
privacy and transaction privacy. Identity privacy publicly
links the real identity and transaction scripts, and there are
several behavioral analysis strategies, including anti-money
laundering and know your customer (KYC) to present the
usage graph. Transaction privacy means the plain contents
on the ledger, including the plain transferring value, account
direction, and so on. Some adversaries may draw attentions
to watch even monitor some accounts with huge amounts of
property. Furthermore, there are several methods to achieve
the high level protected blockchain. [19] employed the
mixer to obfuscate the relationships among people. Maxwell
proposed the Confidential Transaction and firstly achieved
the implemented with the range proof scheme. [20] [21]
finished the privacy preservation protocols based on ring
signature. [22] sealed the plain amounts by using Paillier
cryptosystem. [23] provides a complete solution to make
the sensitive information unreadable for the public. Our
453
system relies on the original chain security and provides
protection on the layer of web servers and back-end. This
design decision comprehensively considers the performance
and security for the whole integrated system as a trade-off.
VII. CONCLUSIONS
In this paper, we presented the ArtChain, which is an
platform designed with registration, tracking, protection, and
provenance for artworks enabled by blockchain technology.
We also discussed how to design, implement and deploy the
blockchain platform in operation as a working product in
practice. The proposed blockchain implementation and Ex-
perimental results showed that our system towards artworks
can provide a complete blockchain-based solution with the
property of irreversibility, authentication, traceability and
transparency.
In future, we plan to work on anti-counterfeiting for the
original works of art by integrating with the smart modules
of IoT, and activating relevant smart hardware and other
functionality (e.g. positioning/location tracking) as required
by artists or collectors.
REFERENCES
[1] Z. Zheng, S. Xie, H.-N. Dai, X. Chen, and H. Wang,
“Blockchain challenges and opportunities: a survey,” Inter-
national Journal of Web and Grid Services, vol. 14, no. 4,
pp. 352–375, 2018.
[2] F. Yiannas, “A new era of food transparency powered by
blockchain,” Innovations: Technology, Governance, Global-
ization, vol. 12, no. 1-2, pp. 46–56, 2018.
[3] Z. Wang, D. Y. Liffman, D. Karunamoorthy, and E. Abebe,
“Distributed ledger technology for document and workflow
management in trade and logistics,” in Proceedings of the 27th
ACM International Conference on Information and Knowl-
edge Management. ACM, 2018, pp. 1895–1898.
[4] M. Zeilinger, “Digital art as monetised graphics: Enforcing
intellectual property on the blockchain,” Philosophy & Tech-
nology, vol. 31, no. 1, pp. 15–41, 2018.
[5] M. McConaghy, G. McMullen, G. Parry, T. McConaghy, and
D. Holtzman, “Visibility and digital art: blockchain as an
ownership layer on the internet,” Strategic Change, vol. 26,
no. 5, pp. 461–470, 2017.
[6] L. Lotti, “Contemporary art, capitalization and the
blockchain: On the autonomy and automation of arts
value,” Finance and Society, vol. 2, no. 2, pp. 96–110, 2016.
[7] A. Whitaker, “Artist as owner not guarantor: The art market
from the artists point of view,” Visual Resources, vol. 34, no.
1-2, pp. 48–64, 2018.
[8] S. De Angelis, L. Aniello, R. Baldoni, F. Lombardi,
A. Margheri, and V. Sassone, “Pbft vs proof-of-authority:
applying the cap theorem to permissioned blockchain,” 2018.
[9] M. Utz, S. Albrecht, T. Zoerner, and J. Str¨
uker, “Blockchain-
based management of shared energy assets using a smart
contract ecosystem,” in International Conference on Business
Information Systems. Springer, 2018, pp. 217–222.
[10] G. Blossey, J. Eisenhardt, and G. Hahn, “Blockchain tech-
nology in supply chain management: An application per-
spective,” in Proceedings of the 52nd Hawaii International
Conference on System Sciences, 2019.
[11] J. Tanner, “Summary of ethereum upgradeable smart con-
tract rd,” https://blog.indorse.io/ethereum-upgradeable-smart-
contract-strategies-456350d0557c, accessed: 2018-11-30.
[12] S. Palladino, “The parity wallet hack explained,” July-
2017.[Online]. Available: https://blog. zeppelin. solutions/on-
the-parity-wallet-multisig-hack-405a8c12e8f7, 2017.
[13] T. Wang, “A unified analytical framework for trustable ma-
chine learning and automation running with blockchain,” in
2018 IEEE International Conference on Big Data (Big Data).
IEEE, 2018, pp. 4974–4983.
[14] K. L. Brousmiche, A. Durand, T. Heno, C. Poulain,
A. Dalmieres, and E. B. Hamida, “Hybrid cryptographic
protocol for secure vehicle data sharing over a consortium
blockchain,” in To appear in Proceedings of IEEE Blockchain
2018, 2018.
[15] H. Zhang, C. Jin, and H. Cui, “A method to predict
the performance and storage of executing contract for
ethereum consortium-blockchain,” in International Confer-
ence on Blockchain. Springer, 2018, pp. 63–74.
[16] I.-C. Lin and T.-C. Liao, “A survey of blockchain security
issues and challenges.” IJ Network Security, vol. 19, no. 5,
pp. 653–659, 2017.
[17] B. Qin, J. Huang, Q. Wang, X. Luo, B. Liang, and W. Shi,
“Cecoin: A decentralized pki mitigating mitm attacks,” Future
Generation Computer Systems, 2017.
[18] M. Wang, Q. Wu, B. Qin, Q. Wang, J. Liu, and Z. Guan,
“Lightweight and manageable digital evidence preservation
system on bitcoin,” Journal of Computer Science and Tech-
nology, vol. 33, no. 3, pp. 568–586, 2018.
[19] D. L. Chaum, “Untraceable electronic mail, return addresses,
and digital pseudonyms,” Communications of the ACM,
vol. 24, no. 2, pp. 84–90, 1981.
[20] S. Noether, A. Mackenzie et al., “Ring confidential transac-
tions,” Ledger, vol. 1, pp. 1–18, 2016.
[21] S.-F. Sun, M. H. Au, J. K. Liu, and T. H. Yuen, “Ringct
2.0: A compact accumulator-based (linkable ring signature)
protocol for blockchain cryptocurrency monero,” in European
Symposium on Research in Computer Security. Springer,
2017, pp. 456–474.
[22] Q. Wang, B. Qin, J. Hu, and F. Xiao, “Preserving transaction
privacy in bitcoin,” Future Generation Computer Systems,
2017.
[23] E. B. Sasson, A. Chiesa, C. Garman, M. Green, I. Miers,
E. Tromer, and M. Virza, “Zerocash: Decentralized anony-
mous payments from bitcoin,” in 2014 IEEE Symposium on
Security and Privacy. IEEE, 2014, pp. 459–474.
454
