Reference 31
Reference 31
Authorized licensed use limited to: Odisha University of Technology and Research. Downloaded on September 18,2024 at 10:16:12 UTC from IEEE Xplore. Restrictions apply.
DNS root 2 P2P network
servers 6 3
7
Authorized licensed use limited to: Odisha University of Technology and Research. Downloaded on September 18,2024 at 10:16:12 UTC from IEEE Xplore. Restrictions apply.
Blockchain technology is DNS/CA based attacks/challenges Legacy solutions Blockchain-based solutions
not feasible for all prob- Difficulty: hard. Immune
DDoS attacks against DNS/CA servers
lems. It is appropriate to Increased number of servers, DDoS mechanisms ... Service: availability
Authorized licensed use limited to: Odisha University of Technology and Research. Downloaded on September 18,2024 at 10:16:12 UTC from IEEE Xplore. Restrictions apply.
Identity management
enables P2P sharing of
personal identities and
related information. It
Global DNS provides greater control
OpenNIC
over personal data and
PKI Mr. Ritchie reduces risks. Identity
verification and a digital ID
can be merged to provide
4
a.b.c.d the functionality of a digi-
2 tal watermark. Blockstack,
Blockchain nodes Web server for example, provides a
P2P network
PKI decentralized public key
distribution system, and
registry for apps and user
1 identities.
3
DNS query https://www.sample.eth
Mrs. Hamilton
FIGURe 3. Blockchain based DNS and DPKI.
applications (dApps) may generate too many tion of subdomains. The legacy DNS system only
transactions on the blockchain, so solutions like allows domain names to be rented for a period of
Plasma propose that each dApp operates on its time. Users will keep their domain names forever
own separate blockchain. with blockchain technology. The users need to
A blockchain-based DNS and DPKI usage sce- deploy a blockchain-based name resolution reg-
nario is given in Fig. 3. There is no need for a CA, istry, implementing a registrar representing a con-
as the public keys are kept in the blockchain as a tract that controls a node. DNSResolver should
DPKI infrastructure. These keys will be used for be set as the domain name resolver with the ser-
the cryptographic functions between the users, vice functions. A user’s NS records can be updat-
Mrs. Hamilton and Mr. Ritchie. Mrs. Hamilton will ed with the user’s registrar.
obtain the digital identity (public key and personal Namecoin was the first blockchain-based
details) of Mr. Ritchie from the P2P network. DNS, which was forked from Bitcoin. Most of
The blockchain-based solutions do not provide the following solutions were derivatives of
the standard domains, so they cannot be thought Namecoin. Namecoin does not seem to be func-
of as a replacement for the legacy system. They tional [11]. KeyId and NXT are more experimen-
are hosting-specific domains like .bit and .eth. The tal naming services, which are more theoretical
domain name database is kept in the ledger, and and not very reliable. Blockstack has its own
all the nodes of that system have the same data- Blockchain Name System. Ethereum Name Ser-
base. Blockchain-based solutions have peering vice is the most valid and functioning system at
agreements with OpenNIC, which aims to be a the moment.
non-national democratic alternative to the tradi- The decentralized namespace should be care-
tional domain registries. OpenNIC provides global fully designed, and it would be a good idea to
DNS namespace and also the blockchain-host- have a hybrid model, where there are also central-
ed domains. OpenNIC servers (https://servers. ized services [11]. As these domains are mostly
opennicproject.org/) are available for public use. used for web browsing, browser add-ons should
The OpenNIC server works in a centralized way, be deployed to reach this decentralized name-
so the system represents a hybrid model, where space.
decentralized blockchains work along with the
centralized solution. Identity ManageMent
The main blockchain-based solutions are clas- Identity management enables P2P sharing of per-
sified in Table 2. Most blockchain-based DNS sonal identities and related information. It pro-
solutions focus on the name resolution. Ethereum vides greater control over personal data and
does have an ID management system, but “Ethe- reduces risks. Identity verification and a digital
reum Name Services” does not. Blockstack and ID can be merged to provide the functionality
DNSchain also provide other services to form a of a digital watermark. Blockstack, for example,
complete solution. The services they provide are provides a decentralized public key distribution
presented in the following sections. system, and a registry for apps and user identities.
The Blockstack application programming inter-
NaMe Resolution face (API) can handle identity and authentication.
The name resolution service is the basic DNS ser- Applications can request permissions from the
vice. Blockchain-based name services provide the users and then gain read-and-write access to the
owner of a domain full control over the distribu- user resources.
Authorized licensed use limited to: Odisha University of Technology and Research. Downloaded on September 18,2024 at 10:16:12 UTC from IEEE Xplore. Restrictions apply.
Identity
tion; it ensures the security with the MITM-proof
Name
management
Distributed Distributed RESTful API.
resolution storage applications
(PKI) Decentralized Internet
Ethereum Name The Internet should be liberated and decentralized
Yes X X X
Services in theory. Violation of net neutrality, censorship,
privacy problems, and disruption of the services
Namecoin Yes Yes X X with DOS attacks are some of the problems we
face today. The standards are not evolving as fast
EmcDNS Emercoin NVC Yes. EMCSSH Yes Yes as they should. DNS records can be censored at
some of the DNS servers, which will cause the
Blockstack Yes. BNS Yes Yes. Gaia Yes domains to be unreachable. DNS is managed as
Yes. Uses Yes. Uses Yes. Yes. Uses
a single and globally unique root. Even though it
DNSChain
Namecoin Namecoin Customizable Blockstack
is managed in a so-called democratic way, being a
centralized solution can be a problem, especially
Table 2. Classification of the blockchain-based solutions. during DDoS attacks.
Decentralized systems can be a robust alter-
native, and can especially replace the central
Distributed Storage systems that need trust. There are several solu-
A decentralized cloud storage (DSN) network tions for the decentralized name resolution ser-
allows sharing and storage of data without hav- vices. Blockchain systems can also be used as a
ing to trust any third parties. This is significant for PKI. Digital certificates and public keys can be
privacy, security, and data control. It also reduc- stored in the ledger. Such a system will not need
es the rate of data failures and outages. DSN is central CAs. The cloud can be used for storage
different from traditional cloud networks with back-ends by implementing the trust issues in a
its client-side encryption, which is more secure decentralized way. Blockstack and Emercoin can
against threats. Proof of retrievability ensures the be given as candidate implementations of such
integrity and availability of the data. The main systems. Cloud security is implemented by proper
advantage of the DSN is flexibility. Speed and selection and careful implementation of the cryp-
low cost advantages can be reached via prop- tographic protocols. The data is written encrypted
er implementation. The biggest concern about on the cloud and then signed.
decentralized storage is the storage capacity. Blockstack proposes a decentralized DNS, PKI,
Keeping the whole ledger in every single node and storage. The authors of Blockstack represent
looks like an illogical solution, especially in it as “the new Internet, where users don’t need to
Internet of Things (IoT) type lightweight devic- trust remote servers” [13]. The implementation of
es. Maintaining the current state of the regis- Blockstack is flexible in such a way that any num-
tered domains and keys is a better idea. There ber of blockchains can be used as communica-
is still need for storage space that is linear with tion channels, and any public cloud can be used
the number of registered domains. It will only for storage. Blockstack also provides a full stack to
require a constant time for checking the integrity build applications for the developers. The system
of the blockchain whenever a new domain is is formed of three components [13]:
registered [12]. • Blockchain: Virtualchain is used to bind infor-
The Gaia storage system is used by Blockstack. mation to public keys and provide trust.
It stores data on behalf of a user after the user • Peer network: The Atlas network is used to
logs into the application. Gaia is used to reuse the provide a scalable index for global data.
existing cloud infrastructure, but writes the data in • Storage system: Gaia is used.
encrypted or signed form [13]. Storj works as a The Emercoin system uses EmcDNS, a PKI ser-
P2P cloud storage network. vice called EmerSSH, and storage. EmcDNS is a
decentralized domain name service, which sup-
Decentralized Applications ports a full range of DNS records of any kind in
name-value format. Emercoin preserves an agree-
DApps are a concept wherein anyone can publish ment with the DNS provider OpenNIC. Users
their apps. Unlike today’s apps, it does not need can reach the domains that are registered with
a third party to gain access to the user’s informa- EmcDNS through the OpenNIC DNS servers. The
tion. The app will remain in its original form as the following can be mentioned as differences from
ledger is immutable. This makes DApps unstop- Blockstack:
pable and resistant to censorship. DApps can be • Emergate.net: This is an experimental work,
developed for money management, e-voting, gov- which will serve as a public gateway to all
ernance systems, and more [14]. The main con- EmcDNS zones by using the URL addresses.
cerns about DApps are [15]: • Emercoin wallet: DNS records can easily be
• The security risks of running anonymous or retrieved from any Emercoin wallet using the
incomplete code three types of user interface, or by the stan-
• Scalability problems dard RFC1034 DNS protocol, which is built
• Resiliency of the application platform in on every Emercoin wallet.
Two DNS DApps examples are Blockstack
and DNSChain. Blockstack is a decentralized Practical Experience
web application, which is in a modular, layered We tried existing tools and applications in our
structure that enables the modules to be con- lab. The Blockstack browser is installed on local
figured with different software. The DNSChain machines for creating (name.id) and managing
system provides simple and secure key distribu- the personal profile associated with it. These
Authorized licensed use limited to: Odisha University of Technology and Research. Downloaded on September 18,2024 at 10:16:12 UTC from IEEE Xplore. Restrictions apply.
identities are then registered on the blockchain. much electricity resources. However, there are
Typical implementations show that this system is other consensus protocols, such as PoS, which The decentralized
used mostly for identity management. Name.id require less resources, and these protocols can be Internet is not a dream.
is associated with the user’s public key, which is enhanced to reach acceptable security at lower Decentralized Infrastruc-
a cryptocurrency wallet address. The profile data cost. The decentralized Internet is not a dream.
is mostly kept in the cloud environments. Social Decentralized infrastructure-related research ture-related research
media addresses are also linked to the profiles. should be more in focus, but there are new work- should be more on the
Personal profiles are expected to be used ing groups like IETF’s Decentralized Internet Infra-
widely when the usage of DApps increases. structure Research Group (DINRG) addressing focus, yet there are new
Blockstack also sponsors an “Ecosystem-Wide, this subject. This area should be studied in depth, working groups like IETF
Universal” Dapp Store (https://app.co/) where and implementations should be enhanced contin- decentralized Internet
several apps are listed. They also created a criteria uously.
list for DApps such as identity, data encryption, infrastructure research
data storage, and software licences. References group (DINRG) on this
We installed a Blockstack core as a blockchain [1] P. Roberts, “Phony SSL Certificates Issued for Google,
node. Blockstack keeps the records as four layers Yahoo, Skype, Others,” Mar. 2011; https://threatpost. subject. This area should
com/phony-ssl-certificates-issued-google-yahoo-skype-oth-
on top of the Bitcoin blockchain. Name queries ers-032311/75061/, accessed June 30, 2018. be studied in depth and
were tested. Blockstack had 524,876 blocks, and [2] D. Fisher, “Final Report on DigiNotar Hack Shows Total implementations should be
the namespace consisted of 77,706 .id names at Compromise of CA Servers,” Oct. 2012; https://threat-
the time of our test. New TLD namespaces (.site, post.com/final-report-diginotar-hack-shows-total-compro- enhanced continuously.
mise-ca-servers-103112/77170/, accessed June 30, 2018.
.media, .device ... etc.) are said to be coming [3] IETF RFC 1034, “Domain Names — Concepts and Facilities,”
soon. P. Mockapetris, The Internet Society; https://tools.ietf.org/
html/rfc1034, 1987.
root@bcrg_testbed:/home/enisk/blockstack# [4] IETF RFC 2826, “IAB Technical Comment on the Unique
DNS Root,” Internet Architecture Board, Network Working
blockstack consensus Group; https://www.ietf.org/rfc/rfc2826.txt, 2000.
{ [5] IETF RFC 5280, “Internet X.509 Public Key Infrastructure
“block_height”: 524876, Certificate and Certificate Revocation List (CRL) Profile”;
“consensus”: “799b9236dc2b7b- https://tools.ietf.org/html/rfc5280, 2008.
[6] IETF RFC 6844, “DNS Certification Authority Authorization
8311ba44f17738ef4b” (CAA) Resource Record,”; ISSN: 2070-1721, 2013; https://
} tools.ietf.org/html/rfc6844, 2013.
[7] Qualys SSL Labs, “SSL Pulse”; https://www.ssllabs.com/ssl-
Results and Conclusion pulse/, accessed June 30, 2018.
[8] A. Kiayias et al., “Ouroboros: A Provably Secure Proof-of-
We are facing the violation of net neutrality, cen- Stake Blockchain Protocol,” Annual Int’l. Cryptology Conf.,
sorship, and privacy problems, which threaten Springer, 2017, pp. 357–88.
the freedom and usability of the Internet. Denial [9] G. Karame, “On the Security and Scalability of Bitcoin’s
of service attacks cause the disruption of many Blockchain,” Proc. 2016 ACM SIGSAC Conf. Computer and
Commun. Security, 2016, pp. 1861–62.
online services. The standards are not evolving [10] K. Wüst, and A. Gervais, “Do You Need a Blockchain?,”
as fast as they should. Decentralized blockchain IACR Cryptology ePrint Archive, 2017, p. 375.
technologies can be developed as a solution. [11] H.A. Kalodner et al., “An Empirical Study of Namecoin and
Blockchain implementations that give name Lessons for Decentralized Namespace Design,” WEIS, 2015.
[12] S. Wilkinson et al., “Storj A Peer-to-Peer Cloud Storage
service and host-specific extensions like .bit Network,” 2014.
and .eth. can also be peered by other services [13] M. Ali et al., “Blockstack: A New Decentralized Internet,”
like OpenNIC. The hybrid solutions are not fully White Paper, 2017.
decentralized, but they are still important and [14] V. Buterin, “A Next-Generation Smart Contract and Decen-
tralized Application Platform,” White Paper, 2014.
serve their purpose. The importance of such a [15] A. Castano, “Why I’m Betting on Blockstack to Save the
solution is that there will not be only a single enti- Decentralized Internet,” Sept. 2017; https://medium.com/@
ty managing the namespace, but also some other alidcastano/why-im-betting-on-blockstack-to-save-the-decen-
alternatives as well. These solutions also work as a tralized-internet-56f65a11fec4, accessed June 30, 2018.
distributed public key infrastructure. The existence
of many nodes on the P2P network serving the Biographies
namespace will serve availability during DDoS Enis Karaarslan ([email protected]) is an assistant pro-
fessor in the Department of Computer Engineering at Mugla
attacks. Sitki Kocman University. He received his Ph.D. in computer
Blockchain-based DNS and PKI implementa- engineering (2008) from Ege University. He was a post-doctoral
tions are not mature enough yet, but the services researcher at EC JRC-IPSC, Italy (2011–2012). He is the head of
they can deliver are promising. There are chal- the MSKU Blockchain Research Group. His research areas are
computer networks, security, privacy, and blockchain. He has
lenges to be solved, such as scalability and energy over 40 papers to his name.
consumption. New solutions, including the Light-
ning network and Plasma, have been proposed E ylul A dıguzel ([email protected]) is an MSKU
for the scalability issues. The blockchain should Blockchain Research Group member who is eager to learn the
fundamentals and technical barriers of this technology. She has
only be used for keeping records; the data should worked on using blockchain effectively for e-voting systems for
be kept in the cloud. Most blockchain implemen- her finishing thesis. She received her B.S. in computer engineer-
tations use PoW consensus protocols and too ing in 2018.
Authorized licensed use limited to: Odisha University of Technology and Research. Downloaded on September 18,2024 at 10:16:12 UTC from IEEE Xplore. Restrictions apply.