A Comparative Analysis of Current Cryptocurrencies
Lara Mauri
1
, Stelvio Cimato
1
and Ernesto Damiani
2
1
Computer Science Department, Università degli Studi di Milano, via Bramante 65, Crema (CR) 26013, Italy
2
EBTIC Laboratory, Khalifa University, Abu Dhabi Campus, PO Box 127788, Abu Dhabi, U.A.E.
Keywords: Cryptographic Protocol, Blockchain, Bitcoin, Ethereum, Ripple.
Abstract: Blockchain technology is having a deep impact on the financial and technical sectors providing a mechanism
for the creation of decentralized currencies and a number of applications in different fields. At the core of the
technology there is a consensus protocol enabling the maintenance of a distributed ledger. In general current
systems are complex schemes that implement a combination of cryptographic algorithm, distributed
techniques, and incentive driven behaviour. In this paper we focus on three of the most diffused platforms,
i.e. Bitcoin, Ripple, and Ethereum, and provide a comparative analysis of their most important features such
as the architecture, the scripting language, the economic and security properties.
1 INTRODUCTION
Digital currencies such as Bitcoin (Nakamoto, 2008)
have pioneered a new approach to tracking financial
transactions. Bitcoin’s breakthrough was to combine
existing techniques (distributed ledgers, public-key
encryption, consensus protocols, Merkle tree
hashing) in an innovative way. The underlying
blockchain architecture is essentially a type of
distributed ledger, i.e. a digital record of ownership
maintained on a distributed network of computers.
The identical copies of the digital ledger (a
continuously growing list of transaction records) are
shared among the network participants and updated
automatically every time a new transaction occurs.
Since additions to the ledger are decided on a
consensus basis by multiple entities, once a
transaction is entered in the blockchain, the record is
immutable and therefore provides an auditable
history of events that cannot be modified. In addition,
blockchain is censorship resistant, which means that
no actor can prevent a legitimate transaction from
being added to the ledger; the integrity of the ledger
is maintained by reaching a consensus about its state.
Ledgers are the basic consensus technology of
transaction at the heart of all modern business models.
The traditional approach to certify the correctness of
transactions requires complete reliance upon a single
centralized authority with formal responsibility over
the system. On the contrary, the blockchain
architecture acts as an alternative value transfer
system that breaks the paradigm of centralized
consensus by using a ledger replicated among
different nodes in a peer-to-peer network. Such
decentralized scheme represents a completely new
means of forming consensus reliably across time.
Moreover, unlike highly centralised systems in which
there is a single point of failure, the blockchain design
does not need to rely on a trusted third party thus
eliminating the risks that come with data being held
centrally. Indeed, the system is designed so that it is
maintained collaboratively by the decentralised
network of nodes that can initiate direct transactions
of data.
Distributed ledgers are highly efficient because
authorized ledger changes are immediately reflected
in all copies of the record. Also, changes by
participants without the necessary permission to
modify the blockchain are rejected, so corrupting the
ledger is extremely difficult. Although distributed
ledger technology (DLT) was created to allow a
global network to securely transact and exchange
value without the need for a third-party organization
in the middle, firms and other institutions are now
actively exploring the practical applications for
blockchain technology beyond the financial world
(Swan, 2015), particularly in the areas of health,
science, government, culture and art. Thanks to their
immutable nature and processing capability, these
digital technologies can be applied to a wide range of
industries and services and thus facilitate the grow of
a sharing economy.
Contribution In this paper, we study the broad field
of DLT. We provide a comparative analysis across
Mauri, L., Cimato, S. and Damiani, E.
A Comparative Analysis of Current Cryptocurrencies .
DOI: 10.5220/0006648801270138
In Proceedings of the 4th International Conference on Information Systems Security and Privacy (ICISSP 2018), pages 127-138
ISBN: 978-989-758-282-0
Copyright © 2018 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
127
the most widely diffused platforms, and specifically,
we scrutinize the design of Bitcoin, Ethereum and
Ripple, with the intent of pointing out the differences
in strategies adopted by them. We devote particular
attention to implementing choices, functionalities and
related concepts. More in details: in Section 2 we
consider the alternatives in types of membership
allowed by the models, along with other components.
Then, in Section 3 we stress the role of blockchain in
the actualization of a structural change in the
economy. Section 4 analyzes the consensus models
used by the different protocols. Section 5 describes
the scripting features, followed by an inspection of
some peculiarities of the platforms in Section 6. A
discussion on some security aspects is presented in
Section 7 while we draw some final conclusions in
Section 8.
2 BLOCKCHAIN BASED
SYSTEMS
The Bitcoin cryptographic currency first popularized
the concept of blockchain. Up until Bitcoin and its
distributed globally-shared ledger was invented,
digital assets were entirely managed by centralized
authorities keeping track of all transactions (BitFury
Group, 2016). In blockchain-based architectures, the
lack of a central entity who exercises concurrency
control over the system has shifted this responsibility
to a whole network providing a new way of mitigating
the risks associated with traditional systems.
Since the emergence of Bitcoin protocol, various
distributed consensus models have been proposed
with novel means of establishing consensus between
parties and divergent areas of focus.
Ethereum is one of the most prominent second-
generation blockchain technology, whose intent is to
create a new model for building decentralized
applications (Buterin, 2014a). The protocol is based
on a combination of virtual machines and stack-based
architecture, featuring a Turing-complete scripting
language, which opens the doors to a hypothetically
unlimited range of applications.
Even though the term “blockchain technology” is
often treated as a synonym for DLT, blockchain
represents one specific type of distributed ledger in
which blocks of transactional data are linked to one
another with data integrity maintained by
cryptography. Consensus ledgers are one of the other
possible structures that can be used to implement
DLT arrangements and Ripple (Schwartz et al, 2014),
in particular, is a type of consensus-oriented
distributed database which targets financial use cases.
This system is not a blockchain technology, since
transactions are not added in the form of blocks that
are hash chained to each other. However, similarly to
the block-oriented consensus models, the Ripple
network updates the ledger through a consensus
process between nodes.
2.1 Models and Structural Features
The decentralized paradigm, on which the blockchain
technology is based, transfers authority and trust to a
network open to all and enables its nodes to transact
with varying degrees of control of the ledger. As the
technology matured, development efforts by financial
institutions have made it possible to adapt this
paradigm to existing contexts in which data is not
completely transparent. Table 1 outlines the types of
membership governing each model, as well as
information related to transaction processing and data
structure used to efficiently verify the integrity of all
the transactions.
Table 1: Compared analysis of the architecture for the three
platforms. Symbols: means that the protocol has the
feature and ‘’ that it has not.
Blockchain-based systems
Bitcoin
Ethereum
Ripple
Permissioned (1) /
Permissionless (2)
(2)
(2)
(1)
Public / Private
Public
Public
Public
Trade history
recorded
Data structure
Merkle
tree
Patricia
tree
Merkle
tree
Notion of account
Multiple senders
and receivers
Depending on whether or not the network nodes are
required to authenticate themselves to participate in
the voting process, two main categories of blockchain
are distinguished: permissionless blockchain and
permissioned blockchain. The former is a distributed
consensus ledger in which no authorization is
required to perform transaction processing. There are
no restrictions on identities of nodes which act as
verifiers and thus anybody can join the network to
become a participant in the consensus process as well
as create blocks of transactions. Bitcoin and
Ethereum are open-ended systems and consequently,
any node can take part in the consensus process to
ICISSP 2018 - 4th International Conference on Information Systems Security and Privacy
128
advance the blockchain. Ripple takes a different
direction and relies on a permissioned setup. Here
only entities belonging to a predefined list of subjects
with known identities can become a verifier in the
system. Indeed, Ripple requires each node to define a
Unique Node List (UNL), which is a set of trusted
authorities. During the process execution, validating
servers vote on the validity of all transactions
received, and each of them consult its list of chosen
validators in order to reach consensus. Since
validators are preselected by an authority, all the
participants in the protocol are known and trusted,
meaning that they are also accountable according to
laws. Permissioned blockchains are based on trust
models and therefore, they better comply with
existing regulations, where only trusted members are
allowed to participate in the consensus; this
configuration maintains compatibility with the
traditional finance setting, but at the cost of reduced
censorship-resistance. Transaction processing is
performed in a more controlled setting, which
operates in an inherently biased way. This exposes the
system to dishonest behaviours by users who might
be involved in fraudulent activities.
Instead, permissionless systems minimize human
factor by replacing it with rigorous approaches, which
are enforced automatically, thereby eliminating trust
involved in interaction between parties and associated
vulnerabilities.
According to (BitFury Group and Garzik, 2015),
there is a second categorization of blockchain types
that is based on the access to the ledger data itself:
public blockchain and private blockchain. In public
blockchains, including Bitcoin and Ethereum, all
records are publicly transparent. Everyone can
participate in the process for determining the next
block to be inserted into the chain as well as read
transactions. This type of blockchain is considered
fully decentralized. At the other end of the spectrum,
private blockchains rely on the model of user rights,
where the possibility to read or modify the blockchain
data is restricted to entities within an organisation.
Since write permissions are kept centralized to one
organization, this configuration could be defined as a
traditional centralized network. However, there are
nuances to the private model concept depending on
the underlying degree of control. For example, since
Ripple is built with a public-based architecture, it is
technically a public system. Nevertheless, it can be
understood as an intermediate model between the
public and the private one because of its closed and
monitored nature. Regardless of the type of allowed
membership, one of the main benefits offered by
distributed ledger technologies is that any asset
referenced in a transaction is traceable through the
blockchain up until its creation. Bitcoin and other
systems which extend on its innovations keep an
updated record of all the transactions that ever
occurred in the history of the network. Ethereum also
has the property that every block contains a copy of
both the transaction list and the current state of its
network. To be more precise, the state is not directly
stored in each block; every block header contains
only the state root, i.e. the root hash of an evolved
Merkle tree, called Patricia tree, that stores the entire
state of the system. These trees record key/value
bindings, where the keys are addresses and the values
consist of account declarations, and boost efficiency
by introducing some modifications to the data
structure (Ethereum Wiki (b)). In addition to the state
root, every block header in Ethereum contains two
other roots related to transactions and receipts
respectively. The property such that each block can
use the root hash of the Patricia tree to refer to the
entire tree is particularly useful when the difference
of state between two blocks is fairly small because
unchanged data of the tree can be simply referenced
without the need to store them twice.
As previously mentioned, Patricia trees are
specialized kind of Merkle trees. The original
application of Merkle proofs introduced in Bitcoin
provides a very efficient process to verify whether a
transaction is part of a block (Antonopoulos, 2014).
Each header block includes the Merkle root, which is
the hash of all the hashes of all the transactions in the
block. The efficiency of Merkle trees is leveraged by
the simplified payment verification (SPV), which
allows light clients to verify that a transaction has
been accepted by the network by downloading just
the chain of block headers, instead of downloading
the entire blockchain. However, since SPV nodes do
not have all transactions and do not download full
blocks, they can prove that a particular transaction is
included in a block, but cannot prove anything about
the current state (Buterin, 2015).
Multi-level data structures represent also an
important scalability feature of the Ripple protocol.
The Ripple ledger chain is similar to the blockchain
architecture. Every ledger is meant as a notional
structure that consists of a previous ledger, a
transaction tree (containing the transactions that have
taken place since the previous ledger) and a state tree
(showing all the account balances, settings, etc.)
(Ripple Labs). So, in contrast to blockchain, each
Ripple ledger contains only the transactions that
created the current ledger from the previous one,
rather than the entire transaction history. This means
that the ledger chain stores all transactions, so it is
A Comparative Analysis of Current Cryptocurrencies
129
possible to rebuild the trade history and confirm the
results of a specific transaction, but the information
that a node needs to store to serve in a fully functional
way is reduced to just the last ledger state.
Ripple users are equipped with a pair of
signing/verification keys to securely send
transactions. Any proposed change to the shared
global ledger, i.e. every transaction, is signed with the
private key of the wallet owner and the corresponding
public key is included in the transaction. The ECDSA
algorithm is the basis for its public key cryptography
and Ripple uses the same elliptic curve specifications
as in Bitcoin except for a different leading byte in the
address format. Unlike Bitcoin and Ethereum, the
current Ripple system only supports transactions
involving a single sender and a single receiver.
Recently, Moreno-Sanchez et al (Moreno-Sanchez et
al, 2017) proposed a protocol called PathJoin,
providing an innovative combination of the
functionality of Ripple protocol and a distributed
threshold signature scheme that overcomes this
limitation and performs atomic transactions.
Instead of explicitly have user identification, the
Bitcoin system uses public key security in which
bitcoins are linked to public keys through unspent
transaction outputs (UTXOs). The UTXO model
requires that each transaction input has a referral to
the output of a previous transaction and the concept
of user balance is a derived construct created by the
cumulative amount of unspent bitcoins associated
with the corresponding public key of the user.
So, the Bitcoin format of transaction implies that the
concepts of accounts and balances are essentially
defined as states of ownership. This peculiarity
cannot be found in the Ethereum ecosystem, in which,
on the contrary, there is a built-in notion of higher-
level concepts such as accounts. In particular,
Ethereum manages two types of accounts: externally
owned accounts (EOAs) and contract accounts.
The externally owned account defines the basic form
of account, which is controlled by a private key and
akin to the abstract concept of Bitcoin accounts.
EOAs have no code and are directly controlled by real
human beings; therefore, the simple ownership of the
private key associated with an EOA gives the ability
to send ether (Ethereum’s intrinsic currency) and
messages from it.
In contrast, contract accounts, on the other hand, are
controlled entirely by code and their code is executed
whenever they receive instructions in the form of a
transaction from an EOA
.
3 ECONOMICS
Unlike traditional payment systems, which typically
involve the transfer of funds denominated in a
sovereign currency, Bitcoin has its own unit of
account called bitcoin. In essence, blockchain-based
assets could be classified as bearer assets, i.e. the
ownership of an asset amounts to nothing more than
the knowledge of the corresponding private key.
Bitcoins are native tokens with no reference to any
underlying commodity and for this reason they have
no intrinsic value (Murphy et al, 2015); they have no
physical form and their supply is not determined by a
central bank. As shown in Table 2, currently, about
17 million BTC are in circulation (CoinMarketCap).
However, the Bitcoin monetary policy is based on
artificial scarcity, i.e. bitcoins are created in such a
way that their numbers are inherently scarce. As it
stands now, the total value of Bitcoin currency units
in circulation represents 79% of the total coins that
will be produced. Indeed, the supply of bitcoins is
programmed to grow at a steady rate and the total
amount of BTC that can be created is capped at 21
million, which is predicted to be reached in
approximately 2140.
Table 2: Compared analysis of the economic features for
the three platforms. Symbols: means that the protocol
has the feature and ‘’ that it has not.
Economics
Bitcoin
Ethereum
Ripple
Native
currency
BTC
ETH
XRP
Circulating
supply
16,684,58
7 BTC
95,779,68
1 ETH
38,622,892,
459 XRP
Maximum
supply
21 million
BTC
100 billion
XRP
Deflationary
nature
The finite monetary base is a clear indication of the
fact that the currency has an inherent tendency to be
deflationary. Since in Ripple the number of XRP
(Ripple’s native currency) starts off at an all-time
maximum of 100 billion, its ecosystem has the same
deflationary behaviour. However, while the Bitcoin
network creates new bitcoins through the mining
process and the number of BTC in circulation
increases more and more slowly as time passes (until
eventually stabilizing at the permanent amount of 21
million), Ripple has no fair mechanism for coin
introduction into the network. As a result, XRP is
ICISSP 2018 - 4th International Conference on Information Systems Security and Privacy
130
designed to be a scarce asset and its supply decreases
each time a transaction fee is paid; to date, there are
about 39 billion XRP in existence available to the
public (CoinMarketCap). Another important
component of the Ripple network to consider is the
fact that it natively supports cross-currency payments,
thereby allowing parties to transact in the currency
they desire. Therefore, it does not only promise to
facilitate the exchange between currencies within its
network, but also aims to be integrated with existing
financial infrastructures, becoming a bridge between
traditional banking and emerging electronic payment
mechanisms (Rosner and Kang, 2016; Swanson,
2015).
Currently there are many industry leaders evaluating
blockchain technology and almost every major
financial market institution in the world is doing DLT
research. Banks and financial market leaders are
adopting blockchain solutions dramatically faster
than initially expected. Two market study reports
from the IBM Institute for Business Value (Bear et
al., 2016a, 2016b) found that in 2017, 14% of finan-
cial markets institutions and 15% of banks expect to
have blockchains in production and at commercial
scale. According to the studies, more than 70% of
these early adopters are prioritizing blockchain
efforts in order to change the role of hierarchies and
trust. By replacing top-down control with distributed
consensus, blockchains are seen as a means to create
new business models and access new markets.
Ethereum (the second platform for market
capitalization, after Bitcoin) represents one of the
most exciting ongoing developments in the
blockchain scenario (Davidson et al, 2016). Ether
(Ethereum’s value token), is meant as a system
resource needed to execute smart contracts, i.e. the
mechanism carrying out the direct exchange of values
between untrusted parties. Unlike the other two
protocols, Ethereum has infinite monetary base,
meaning there is no hard limit for this particular coin;
the issuance of ether occurs at a constant annual linear
rate via the mining process, but the rate of growth of
the supply is not constant. The amount of ether
available at the present time and circulating in the
market is roughly 96 million (CoinMarketCap).
4 CONSENSUS MECHANISMS
Blockchain technology enables entities to transact
directly even though they do not trust each other,
without the need to send proposed transactions to a
centralized third-party acting as a trusted
intermediary. In contrast to traditional centralized
systems, participants collectively maintain the
common shared ledger in a decentralized fashion
allowing a faster reconciliation between transacting
parties. Each node keeps an individual copy of the
blockchain, whose state is updated as newly-mined
blocks are added. However, nodes may have
completely inconsistent view of data recorded on the
blockchain due to the divergent order in which
transactions are listed in the replicas. Therefore,
participants need to coordinate among each other in
order to determine the legitimate ledger and guarantee
the consistency of the system across all nodes.
Achieving consensus in a distributed setting is not a
trivial task due to the strong dynamism of the system
and the numerous actors involved. In addition, the
network has to be able to operate properly even in the
presence of malicious nodes that can cause Byzantine
faults. These failures were first identified and
described by (Lamport et al., 1982) as the Byzantine
General’s Problem. The original problem description
characterizes the case of a group of generals, each
commanding a portion of an army, trying to formulate
a common plan of action for attacking the surrounded
enemy city. Generals communicate only through
messengers to reach a mutual agreement. However,
some generals might be traitors and hence have the
potential motivation to lie and distort messages (such
an adversarial environment is comparable to the
situation of a distributed system).
From a general perspective, one of Bitcoin’s main
contribution is a solution to this problem by means of
a consensus mechanism that safeguards the stability
of the network. The consensus method enforces a
temporal and unambiguous ordering of transactions
in the system, thereby ensuring a unique authoritative
view of the world state. Proof-of-Work (PoW) is the
mechanism introduced by Bitcoin to reach a
consensus on the distributed shared state. In essence,
the algorithm requires each node to show that it has
performed some amount of work in order for a set of
new transactions (block) to be accepted by the
network members. The PoW consists in the search for
a specific hash digest that is less than a certain
difficulty level fixed by the system. The idea behind
proof-of-work is simple and involves repeatedly
hashing the header of the newly formed block, which
the participant is wishing to publish, together with a
nonce until a solution with desirable properties
emerges; practically, the goal is to find a hash starting
with a certain number of zero-bits. Since the next
block always includes the PoW hash of the last mined
block, the structure resulting from the consensus
process is a hash chain which grows incrementally
(hence the name of blockchain).
A Comparative Analysis of Current Cryptocurrencies
131
Bitcoin’s PoW scheme laid the theoretical foundation
for modern consensus models and new ones are still
emerging providing innovative functions and
properties (Cachin and Vukolić, 2017). As Table 3
outlines, Ethereum currently proposes Ethash as its
specific proof-of-work algorithm (Ethereum Wiki
(a)). It requires to randomly create a dataset
(initialized by the current blockchain length), which
forms a DAG (directed acyclic graph) and then
attempt to solve a specific constraint on it. Each
Ethereum client has to generate the DAG, which is
regenerated every epoch, i.e. every 30000 blocks.
Table 3: Compared analysis of the consensus models for the
three platforms. Symbols: means that the protocol has
the feature and ‘’ that it has not.
Bitcoin
Ethereum
Ripple
Verification
method
PoW
PoW
(Ethash)
Consensus
Involved nodes
All
All
Validators
Hash algorithm
SHA-
256
Keccak-
256
SHA-
512Half
Bock time
10 min
~14 sec
5-10 sec
Transaction rate
Low
Low
High
Probabilistic
transaction
finality
Mining reward
Transaction fee
Although the proof used in Ethereum is similar to
Bitcoin PoW protocol, the point in Ethash is that it
uses a different cryptographic primitive for its
hashing function, called Keccak-256 (a variant of the
FIPS 202 based standard (SHA-3) (FIPS 202, 2015),
instead of relying on the double SHA-256 hash
algorithm.
Proof-of-work represents the technique commonly
used for achieving fault tolerance in a distributed
system. However, there are models that use a
different approach to demonstrate consistency of
transactions. Among them, Ripple is a current
application of Byzantine agreement which opted for
a voting scheme in order to advance the state of the
shared ledger. It implements a round-based process
called consensus, whose objective is to make it
possible for all the nodes to agree on which
transactions to include in the last closure of the
ledger. A crucial aspect of this protocol is the fact that
only a group of nodes can participate in advancing the
ledger, thus acting as validating servers, and the
system is based on trust relationships, since each
server only considers the proposals from the nodes in
its UNL. In an iterative process, validators vote on the
validity of transactions received and the ledger is
finalized when a super-majority of 80% of votes from
servers in the UNL is reached. After a consensus
round completes, validating servers calculate a new
version of the ledger and transmit their results to the
network. In particular, they publish a signed hash of
the ledger, called validation, whose purpose is to
ensure that all participants derive the same ledger.
The hash function used in Ripple is known as SHA-
512Half. It provides a hash value that is calculated by
applying SHA-512 to some contents; then the result
is truncated to the first 256 bytes. Since in Ripple
there is no process of solving the proof-of-work
cryptographic puzzle, consensus is fast and ledgers
are validated in seconds. As a new ledger is closed
approximately every 5 seconds, Ripple takes on
average from 5 to 10 seconds to confirm a transaction.
This is significantly different from Bitcoin block
time. Indeed, a new block is generated roughly every
10 minutes. The Ethereum network, by comparison,
produces a block every 14 seconds on average.
One of the important factors that requires close
attention when evaluating a blockchain platform is
perhaps the transaction finality (Buterin, 2016). This
criterion indicates whether a transaction included in
the blockchain is truly finalized, i.e. there is no way
to revert that operation. Typically, decentralized
systems provide this property probabilistically.
Because blocks are calculated in a distributed setting,
two independent nodes can discover a new block and
broadcast it almost simultaneously. This leads to the
creation of a temporary fork, an occurrence that can
be exploited by a malicious node to intentionally
cause a reorganization in the ordering of transactions.
Even though Bitcoin solves this problem by requiring
that users accept the longest branch (i.e. the branch
involving the highest amount of computational effort)
as the true one, in practice, there is no system that
offers truly complete finality. However, a user can
consider a transaction practically final by means of
the number of confirmations, which represent the
blocks depth. In the specific case of Bitcoin,
transactions are confirmed after the generation of six
consecutive blocks and hence, a payment is
confirmed after one hour on average. As a result, in
systems operating in the same way as Bitcoin,
transaction finality is a probabilistic concept that is
heavily dependent on the waiting time required by a
transaction before being considered as confirmed and
then final. Both Bitcoin and Ethereum are incapable
of dealing with high transaction rates because of such
ICISSP 2018 - 4th International Conference on Information Systems Security and Privacy
132
a slow transaction confirmation mechanism. Ripple,
conversely, is a model designed for high performance
and immediate transaction finality, since once a
transaction is included in the last validated ledger, it
is confirmed fast and cannot be reversed.
An important function in maintaining the
immutability of the blockchain system is that of
incentives. As it is known, DLT works through means
of game theory, since the consensus process requires
co-operation between actors with unaligned interests.
The game-theoretic approach plays a critical role in
achieving a balanced strategy to maintain a unified
version of the global ledger. As such, PoW-based
systems cannot function without economic rewards
for nodes participating in the creation of new blocks
(BitFury Group, 2015). Bitcoin incorporates
incentive mechanisms, which come in the form of
mining rewards and transaction fees. The first node
that successfully solves the PoW and gets to add its
block to the blockchain can collect the block reward.
Currently, it amounts to 12.5 bitcoins. This issuance
is determined algorithmically: the reward halves
every 210,000 blocks (roughly every four years). The
next halving is expected to occur in June 2020
(Bitcoin Clock). Ethereum introduced a similar
wealth distribution mechanism by which the
successful PoW miner receives a static block reward
of 5 ETH. The beneficiary also receives the gas
expended by executing the transactions in the block,
where gas is the fundamental network cost unit at the
base of each transaction. Moreover, there is an extra
reward equal to 7/8 of the static block reward for
including uncles (i.e. stales blocks), with a maximum
of two uncles allowed per block. The approach
adopted by Ripple differs completely from the models
just described as it does not provide a direct monetary
reward for nodes support. The reason behind such
lack of an incentive mechanism lies in the fact that the
XRP were premined and the protocol does not have a
mining process.
On the other hand, the second Bitcoin revenue stream
is represented by the voluntary transactions fees
associated with a block. If the value of a transaction
input exceeds the value of the output, the net
difference in value may be claimed as a transaction
fee, which serves as an incentive to make sure that a
particular transaction will get included into the next
block. Also in Ethereum’s environment, the variable
fee is a way to prioritize some transactions over
others. Fees in Ethereum are denominated in gas,
which represents the amount of ether that covers the
cost of executing a transaction. So every transaction
must contain, alongside its other data, two further gas
related fields: a gasprice value and a startprice value.
The latter defines the maximum amount of gas that
the transaction sender is willing to pay. Ether is used
to purchase gas, which is bought prior to the
execution of a transaction at a certain price. At the end
of the transaction, if not all the gas is consumed,
unused gas is refunded in ether to the sender’s
account; thus, ether is converted freely to and from
gas as required. Instead, if the gas runs out while the
transaction is being executed, this is treated as an
exception: the state is completely reverted, but the
ether used to purchase the gas is not refunded.
In Ripple each transaction submitted to the network
requires a transaction fee specified in XRP. This cost
is designed to increase based upon the transaction
load (currently, the minimum transaction cost is
0.00001 XRP). The novelty associated with this kind
of fee is the fact that, unlike the other two protocols,
there are no beneficiaries and therefore, it is not paid
to any party. Once a transaction is included in a
validated ledger, the exact amount of XRP specified
by the fee parameter is irrevocably destroyed.
Consequently, this approach creates a sort of artificial
scarcity that makes XRP more valuable and drives a
tendency toward the concentration of wealth.
5 SCRIPTING
The Bitcoin UTXO model characterizes how
transactions move value from transaction inputs to
transaction outputs. Since bitcoins are thought of as
unspent outputs associated to a public key, transfers
occur in a chain of transactions consuming and
creating UTXO. Each input of a transaction refers to
a given previous output, which is defined by a script.
So transaction validation is achieved through the
execution of a scripting language, whose goal is to
ascertain under which condition is it possible for a
user to spend the outputs (Table 4).
Table 4: Compared analysis of the scripting features for the
three platforms. Symbols: means that the protocol has
the feature and ‘’ that it has not.
Scripting
Bitcoin
Ethereum
Ripple
Built-in script
language
Smart contract
implementation
Turing-completeness
A Comparative Analysis of Current Cryptocurrencies
133
Specifically, Bitcoin has a locking script which
specifies the spending conditions and an unlocking
script which satisfies such requirement and allows the
output to be spent (Bonneau et al., 2015).
However, the scripting functionalities of the
underlying technology behind Bitcoin can be used for
application domains that go beyond currency. For
example, the expressiveness of the blockchain
technology can be used to implement what is known
as smart contract, a concept introduced by Nick
Szabo in 1997 (Szabo, 1997). Smart contracts
represent a technical advancement to the practice of
law, which formalize the contractual obligations into
programming code and verify them in a self-
executing way, eliminating ambiguity problems of
natural languages. Since contracting parties can
structure their relationships without involving a
trusted third party, the decentralized scheme of math-
based currency systems like Bitcoin, Ethereum and
Ripple makes distributed ledger networks suitable for
smart contracts. In 2014, Ripple released the first
prototype for Codius, a platform that uses the concept
of smart oracles (Schwartz and Thomas, 2014) to
implement smart contracts. However, the
development of the project was discontinued nearly a
year later (Thomas, 2015). Instead, Ethereum is
specifically designed as a multipurpose platform,
featuring smart contract functionality. One of the key
points of this system is that the protocol implements
a completely programmable blockchain enabling the
execution of an unlimited variety of user-
customizable smart contracts. Because of its
resilience to tampering, these computer programs can
be leveraged to automatically conduct transactions or
perform specific actions on the Ethereum blockchain.
The three protocols differ markedly in terms of
scripting: Ethereum’s programming language has a
Turing-complete nature, Bitcoin has several
limitations, including non-Turing-completeness, and
Ripple does not even have a scripting language. The
Bitcoin network has purposefully omitted Turing-
completeness, meaning that the scripts do not support
all computation (e.g. iterative loops). This restriction
has the function to avoid undesired behaviours related
to a problem in computer science known as the
halting problem, according to which there is no
possible mathematical way to determine whether a
computer program will halt or continue to run forever.
Despite the fact that Bitcoin’s scripting is realized by
a simple stack-based language supporting basic
operations, the protocol implements a weak version
of smart contracts. Thanks to an unlimited ability to
implement rich logic, Ethereum takes the scripting
features of Bitcoin blockchain to the next level. The
protocol is designed to execute code of arbitrary
algorithmic complexity, created for any purpose users
deem necessary, where the only limit is given by the
imagination of its developer and associated resources.
User-defined operations need to be interpreted and
this task is handled by the Ethereum Virtual Machine,
which constitutes the key part of the execution model
(Wood, 2014).
6 PLATFORMS PECULIARITIES
As Table 5 illustrates, the platforms have some
differences in the technical features and peculiarities
that derive from the different types of consensus
mechanism they adopt.
Table 5: Compared analysis of the platforms peculiarities.
Symbols: means that the protocol has the feature and
’ that it has not.
Platforms peculiarities
Bitcoin
Ethereum
Ripple
Determinism
Energy
consumption
Wasteful
Wasteful
Efficient
ASIC resistance
At the core, DLT creates opportunities of leveraging
consensus-oriented models for transaction validation
within the context of a distributed ecosystem. A
consensus mechanism is the way in which the nodes
agree on the value of a proposed change that then
updates the shared ledger, and can be described as the
set of rules and procedures that ensures consistency
and authenticity. Following this idea, the logic for
distributed ledger transaction processing must
be deterministic because each participant in the
consensus process has to be able to find the same
result when verifying a transaction. Otherwise, each
node would produce different outputs causing
consensus failure.
Ripple makes extensive use of the concept of
determinism, especially when servers communicate
with each other to agree upon the finality of the state
after processing the transactions as a unit. Indeed,
since all nodes must publish precisely the same
ledger, it is imperative that they converge on the same
transaction set. If there are multiple conflicting
transactions in the same round, participants can sort
and execute them in a deterministic way following
pre-defined rules.
ICISSP 2018 - 4th International Conference on Information Systems Security and Privacy
134
Bitcoin requires nodes to solve a cryptographic
puzzle, which is computationally hard by design, in
order to add a new block to the blockchain. The PoW-
based consensus forces participants to expend
computational resources toward the purpose of
ensuring the safety of the network. Since this effort
has a measurable cost associated to it, Bitcoin relies
on incentive mechanisms to induce nodes to solve the
proof-of-work. Therefore, the PoW concept as used
in Bitcoin comes along with a few drawbacks.
First of all, due to the need for enormous amounts of
computational resources, the protocol consumes a lot
of electricity, and hence it is wasteful in terms of
energy expenditure (O’Dwyer and Malone, 2014).
Secondly, since mining profitability depends on
factors such as hash rate and cost of electricity, the
competitive nature of the process has led miners to
develop more powerful and cost-efficient customized
hardware. As a result, the mining ecosystem has come
to be dominated by ASICs. These are a kind of
integrated circuit specifically built for a certain
purpose (in Bitcoin’s case the task consists in
computing uniquely the SHA-256 hash function),
whose power is to achieve massive gains through
parallelization. With the introduction of ASIC
implementations, the probability of being the first to
successfully find a valid nonce has become very low
for a regular user. Consequently, in order to increase
their chance of winning the mining competition,
participants rely on mining pools, which are groups
of miners that contribute to perform the block
validation jointly; in case of win they split the mining
reward according the contributed processing power.
Today, the majority of bitcoin mining is done in data
centers, by large powerful companies that take
control of the mining power. The emergence of
mining pools, therefore, has caused a departure from
the original Bitcoin idea of decentralization
(Nakamoto, 2008).
By using Ethash, Ethereum combats mining
centralization (a problem that does not affect Ripple
ecosystem, since it does not require mining to achieve
consensus). Its mining algorithm is meant to be
ASIC-resistant, meaning that ASICs are no more
efficient than general-purpose computers at mining.
Ethash is designed to be memory hardness, a property
where the time to compute a valid PoW derives from
the amount of memory required to hold data. The
function takes a very large amount of memory and
this reduces the power of specialized hardware
solutions, thereby encouraging individuals to use
their GPUs and allowing a tangible decentralized
mining process. However, the consensus process is
based on the proof that a particular amount of
computation has been expended in finding a value
less than a pre-defined target threshold and hence,
also Ethereum wastes a lot of energy. Conversely,
Ripple is more environment-friendly thanks to the
absence of mining. Indeed, the energy cost of its
consensus process only derives from the processing
power needed to update the ledger and verify the
transactions.
In light of the energy expenditure, one of the
proposed improvements to the current design of
Ethereum involves the update of the PoW scheme
(Buterin, 2014b). The idea is to move to Proof-of-
Stake (PoS), i.e. an approach in which the probability
to create a block is proportional to a user’s ownership
of cryptocurrency in the blockchain system. The
algorithm is designed to overcome the downsides of
Bitcoin-like PoW mining process by requiring all
nodes to compete with their node fraction of stake. So
it is based on the distribution of digital currency
within the system rather than computational
resources. The proposal for this new scheme, known
as Casper, is still in testing, but a rough
implementation guide has recently been released
(Ethereum Research).
7 SECURITY ASPECTS
By design, once a transaction is added to the
blockchain and confirmed, it can never be reversed.
The existence of multiple shared copies of the same
ledger makes these systems inherently harder to
attack than centralized solutions. However, even
though data integrity is one of the key point of
blockchains, distributed ledgers are not invulnerable
to cyber-attack because legitimate changes to the
global record can be made in principle by anyone.
Table 6: Compared analysis of the security aspects for the
three platforms. Symbols: means that the protocol is
resistant in principle to the attack and ‘’ that it is not.
Security
Bitcoin
Ethereum
Ripple
Double-spending
attack
Sybil attack
51% attack
DoS attack
The double-spending problem refers to a failure case
of digital cash schemes where two separate
transactions sent into the network spend the same
A Comparative Analysis of Current Cryptocurrencies
135
digital currency (a double spend is hence a deliberate
fork). Bitcoin solves this problem by chronologically
ordering blocks of transactions into an ongoing chain
of proof-of-work that is visible to all users. Therefore,
by implementing a confirmation mechanism and
imposing the rule that bitcoins can be spent only from
UTXO, Bitcoin naturally defends against it.
Similarly, Ethereum uses its PoW-based algorithm to
prevent the risk that a user could concurrently spend
the same unit of currency in several transactions.
However, the previously mentioned Bitcoin’s SPV
method, is not always effective in resisting double-
spending attacks. In (Karame et al, 2012), authors
investigate the problem and demonstrate that these
attacks can be performed in spite of the measures
recommended by Bitcoin developers.
Ripple offers an alternative solution to the double-
spending problem through its consensus process. It is
resistant to the attack because transactions are
considered validated only when an overwhelming
majority of validators sign validations for the same
ledger. Since the process requires agreeing on an
order for the transactions, if two transactions are a
double spend, the attack is solved simply by agreeing
on which of the two transactions comes first (the other
is considered invalid, and hence not applied).
As shown in Table 6, all three platforms are immune
to Sybil attacks (Douceur, 2002). Reputation-based
systems are susceptible to this type of attack, in which
a malicious agent assumes multiple pseudonymous
identities with the goal to gain a disproportionately
large influence. By acquiring control over a
substantial fraction of the system, the adversary is
able to manipulate the voting outcomes.
Bitcoin and Ethereum prevent this attack by means of
PoW. The ability of block generation is proportional
to the computing power, and not to the number of
counterfeit identities (Tschorsch and Scheuermann,
2015). Therefore, the capability of the attacker is
determined by the number of blocks it can produce.
The way in which Ripple avoids Sybil attacks is based
on a strategy that is conceptually opposed to the
approach adopted by Bitcoin, because Ripple does
not work in an open environment where anyone can
participate in the consensus process. Indeed, given
that it implements a permissioned system, a Sybil-
resistant strategy is superfluous. Instead of evaluating
proposals from all validators in the network, a
validating server needs to query only its UNL. The
sole fact that consensus is reached by means of trusted
relationships guarantees Sybil resistance.
In the current landscape, blockchain design is
inherently vulnerable to a 51% attack (Swan, 2015).
Blockchain paradigms rely on the assumption that the
majority of nodes act honestly, but in principle it
would be possible for a single malevolent node or
organization to amass a large amount of
computational power and disrupt the stability of the
process. In particular, if an attacker controls >50% of
the mining power, he can create an independent
branch maintaining a fork. Thus if this attack is
successful, a malicious actor can manipulate the
ledger to his advantage and double spend his own
currency, thereby rewriting blockchain history.
Empirical evidence shows that this attack is infeasible
for any single user, as it would require enormous
computational power to recompute all proofs for all
the previous blocks in the chain. However, although
Bitcoin itself is purely decentralized, the declining
incentive to mine leads to centralization of the mining
function. (Beikverdi and Song, 2015) argue that this
trend toward centralization increases the risk of a
51% attack. Another research (Eyal and Sirer, 2014)
proves that the selfish attack (i.e. a method by which
a coordinated group of nodes increase their returns by
not publishing a valid solution to the rest of the
network), facilitates the grow in size of the colluding
group because honest miners strategically decide to
join the selfish miners.
Ethereum shares the same weaknesses as Bitcoin
about the 51% attack, but its ASIC-resistant design
makes the protocol more resistant to the attack.
Ripple, as it is known, does not rely on distributed
computational power to protect the integrity of the
network. It replaces the vote per computing power of
the miners notion of PoW based consensus
mechanisms with the vote per validator (Karame and
Androulaki, 2016). The underlying assumption is that
the majority of Ripple nodes will not collude to
manipulate the voting result. Actually, if 80% of
validating servers collude, it is possible to confirm a
fraudulent transaction. Thereby, the Ripple
equivalent of Bitcoin’s 51% attack is when a group
gets control over a sufficient number of validators to
cause consensus fail. As participants specifically
select their own validating servers, the probability of
success of the attack is very low. However, in the
event that the majority of validators become
malicious, they can rewrite the entire system
transaction history (Armknecht et al., 2015).
Finally, since distributed ledger technologies are
based around a public ledger of information
maintained by a network of computers around the
world, adversaries could broadcast large amounts of
transaction spam in an attempt to disrupt the normal
operation of the network. The most visible
consequence of such attack is the creation of excess
load on the network, which causes difficulties in
ICISSP 2018 - 4th International Conference on Information Systems Security and Privacy
136
processing legitimate transactions. To mitigate
denial-of-service (DoS) attacks, Ripple enforces a
transaction fee. Moreover, unlike other architectures,
Ripple requires each account to have a small reserve
of XRP for the creation of ledger entries in order to
protect the network from abusive creation of ledger
spam. The current minimum amount needed to fund
a new address is 20 XRP. This creates a strong
disincentive against ledger spamming because any
attacks aimed at wasting network bandwidth become
very expensive for malicious agents.
A denial-of-service attack on Ethereum blockchain
would imply that a malicious actor utilizes the
network resources improperly in an attempt to
interfere with the miners’ ability to quickly settle
legitimate transactions. Due to the Turing-complete
nature of the contracts, Ethereum is inherently
vulnerable to DoS attacks, as an attacker could
perform a successful attack by sending transactions
that loop forever. However, it protects the network
against DoS attacks through the use of gas. Gas limits
the number of computational step a transaction
execution is allowed to take, and hence ensures that
there can be no infinite loop. Similarly, Bitcoin has
some DoS prevention built-in. Theoretically, it is
immune to hostile infinite loops because its scripting
language is non-Turing-complete. Actually, Bitcoin
is vulnerable to DoS attacks (Vasek et al, 2014).
8 CONCLUSIONS
In this paper, we presented a survey specifically
targeting the distinguishing features of three of the
most diffused platforms, i.e. Bitcoin, Ethereum and
Ripple. The compared analysis of these systems
focused on their common points, as well as
differences in how they maintain the integrity of data
recorded on the shared ledger, by grouping them into
six main categories. The work includes an accurate
description of the different consensus and incentive
mechanisms adopted by the platforms for securing the
network. Also we examined the scripting features,
security aspects and impact on the economy, as well
as related concepts.
Whereas blockchain is still in its emergent and
immature technological phase, the increasing interest
on it is showing the importance and awareness of
distributed ledgers as one of the most promising
technology that will have a pervasive impact on the
future of many sectors of our socio-economic
systems. Indeed, the emergence of the blockchain
technology could give rise to the next generation
beyond the internet, potentially leading to the creation
of new types of economies. Blockchain’s ability to
catalyse transparency is based on the way it leverages
a global peer-to-peer network to guarantee integrity
of value exchanged between parties without the need
for central authorities. Therefore, by providing a way
of recording transactions securely, distributed ledger
technology offers the opportunity to reimage how the
financial system can work. However, it should not be
understood only as a disruptive technology, but also
as a foundational technology that offers the
possibility to create new foundations for the social
infrastructure (Iansiti and Lakhani, 2017).
REFERENCES
Antonopoulos, A. M. 2014. Mastering Bitcoin: Unlocking
Digital Crypto-Currencies. O'Reilly Media,
Sebastopol.
Armknecht, F., Karame, G. O., Mandal, A., Youssef, F. and
Zenner, E. 2015. Ripple: Overview and Outlook. In:
Proceedings of International Conference on Trust &
Trustworthy Computing, 9229:163180.
Bear, K., Drury, N., Korsten, P., Pureswaran, V., Wallis, J.
and Wagle, L. 2016a. Blockchain rewires financial
markets: Trailblazers take the lead. Executive Report -
IBM Institute for Business Value, [online] Available:
https://www-01.ibm.com/common/ssi/cgi-bin/ssialias?
htmlfid=GBP03469USEN&
Bear, K., Drury, N., Korsten, P., Pureswaran, V., Wallis, J.
and Wagle, L. 2016b. Leading the pack in blockchain
banking: Trailblazers set the pace. Executive Report -
IBM Institute for Business Value, [online] Available:
https://www-01.ibm.com/common/ssi/cgi-
bin/ssialias?htmlfid=GBP03467USEN.
Beikverdi, A. and Song, JS. 2015. Trend of centralization
in Bitcoin’s distributed network. In: Software
Engineering, Artificial Intelligence, Networking and
Parallel/Distributed Computing (SNPD), 2015 16th
IEEE/ACIS International Conference on, pp. 16.
Bitcoin Clock, [online] Available: http://bitcoinclock.com/
BitFury Group 2016. Digital Assets on Public Blockchains.
White Paper, [online] Available: http://www.the-block
chain.com/docs/bitfury-digital_assets_on_public_bloc
kchains.pdf.
BitFury Group 2015. Incentive Mechanisms for Securing
the Bitcoin Blockchain. White Paper, [online]
Available: http://bitfury.com/content/5-white-papers-
research/bitfury-incentive_mechanisms_for_securing_
the_bitcoin_blockchain-1.pdf.
BitFury Group and Garzik, J. 2015. Public versus Private
Blockchains - Part 1: Permissioned Blockchains. White
Paper, [online] Available: http://bitfury.com/content/5-
white-papers-research/public-vs-private-pt1-1.pdf.
Bonneau, J., Miller, A., Clark, J., Narayanan, A., Kroll, J.
A. and Felten, E. W. 2015. SoK: Research Perspectives
and Challenges for Bitcoin and Cryptocurrencies. In:
A Comparative Analysis of Current Cryptocurrencies
137
Proceedings of the 36th IEEE Symposium on Security
and Privacy (SP’15), pp. 104121.
Buterin, V. 2014a. A Next-Generation Smart Contract and
Decentralized Application Platform. White Paper,
[online] Available: https://github.com/ethereum/wiki/
wiki/White-Paper.
Buterin, V. 2014b. On Stake, [online] Available:
https://blog.ethereum.org/2014/07/05/stake/
Buterin, V. 2015. Merkling in Ethereum, [online]
Available: https://blog.ethereum.org/2015/11/15/merkl
ing-in-ethereum/
Buterin, V. 2016. On Settlement Finality, [online]
Available: https://blog.ethereum.org/2016/05/09/on-
settlement-finality/
Cachin C. and Vukolić, M. 2017. Blockchain Consensus
Protocols in the Wild. Technical Report
arXiv:1707.01873, IBM Research Zurich, [online]
Available: https://arxiv.org/abs/1707.01873.
CoinMarketCap, Cryptocurrency Market Capitalizations,
[online] Available: https://coinmarketcap.com/
Davidson, S., De Filippi, P. and Potts, J. 2016. Economics
of Blockchain, [online] Available: http://dx.doi.org/
10.2139/ssrn.2744751.
Douceur, J. R. 2002. The Sybil Attack. In: Proceedings of
the 1
st
International Workshop on Peer-to-Peer Systems
(IPTPS ’02).
Ethereum Wiki (a), Ethash, [online] Available:
https://github.com/ethereum/wiki/wiki/Ethash.
Ethereum Wiki (b), Patricia Tree, [online] Available:
https://github.com/ethereum/wiki/wiki/Patricia-Tree.
Ethereum Research, Casper Version 1 Implementation
Guide, [online] Available: https://github.com/
ethereum/research/wiki/Casper-Version-1-
Implementation-Guide.
Eyal I. and Sirer, E. G. 2014. Majority Is Not Enough:
Bitcoin Mining Is Vulnerable. In: Financial
Cryptography and Data Security, 8437:436454.
FIPS 202, 2015. SHA-3 Standard: Permutation-Based
Hash and Extendable-Output Functions, [online]
Available: http://nvlpubs.nist.gov/nistpubs/FIPS/NIST.
FIPS.202.pdf.
Iansiti M. and Lakhani, K. R. 2017. ‘The Truth about
Blockchain’, Harvard Business Review 95(1):118127.
Karame G. O. and Androulaki, E. 2016. Bitcoin and
Blockchain Security. Artech House, Norwood.
Karame, G. O., Androulaki, E. and Capkun, S. 2012.
Double-spending fast payments in Bitcoin. In:
Proceedings of the 2012 ACM Conference on Computer
and Communication Security, pp. 906917.
Lamport, L., Shostak, R. E. and Pease, M. 1982. The
Byzantine Generals Problem. ACM Transactions on
Programming Languages and Systems (TOPLAS),
4(3):382401.
Moreno-Sanchez, P., Ruffing, T. and Kate, A. 2017.
PathShuffle: Mixing Credit Paths for Anonymous
Transactions in Ripple. In: Proceedings on Privacy
Enhancing Technologies, 2017(3):107126.
Murphy, E. V., Murphy, M. M. and Seitzinger, M. V. 2015.
Bitcoin: Questions, Answers, and Analysis of Legal
Issues, Congressional Research Service, [online]
Available: https://fas.org/sgp/crs/misc/R43339.pdf.
Nakamoto, S. 2008. Bitcoin: A peer-to-peer electronic cash
system, [online] Available: https://bitcoin.org/bitcoin.
pdf.
O’Dwyer, K. J. and Malone, D. 2014. Bitcoin mining and
its energy footprint. In: 25th IET Irish Signals and
Systems Conference and China-Ireland International
Conference on Information and Communications
Technologies (ISSC 2014/CIICT 2014), pp. 280285.
Ripple Labs, The Ledger, [online] Available:
https://ripple.com/build/ledger-format/#tree-format.
Rosner, M. T. and Kang, A. 2016. ‘Understanding and
Regulating Twenty-First Century Payment Systems:
The Ripple Case Study’, Michigan Law Review. 114(4):
649681.
Schwartz E. and Thomas S. 2014. Smart Oracles: A Simple,
Powerful Approach to Smart Contracts, [online]
Available: https://github.com/codius/codius/wiki/Smar
t-Oracles:-A-Simple,-Powerful-Approach-to-Smart-
Contracts.
Schwartz, D., Youngs, N. and Britto, A. 2014. The Ripple
Protocol Consensus Algorithm. White Paper, [online]
Available: https://ripple.com/files/ripple_consensus_w
hitepaper.pdf.
Swan, M. 2015. Blockchain: Blueprint for a New Economy.
O’Reilly Media, Sebastopol.
Swanson, T. 2015. Consensus-as-a-service: a brief report
on the emergence of permissioned, distributed ledger
systems, [online] Available: http://www.ofnumbers.
com/wp-content/uploads/2015/04/Permissioned-distrib
uted-ledgers.pdf.
Szabo, N. 1997. ‘Formalizing and securing relationships on
public networks’, First Monday 2(9).
Thomas, S. 2015. Codius - One Year Later, [online]
Available: https://codius.org/blog/codius-one-year-
later/
Tschorsch, F. and Scheuermann, B. 2015. Bitcoin and
beyond: A technical survey on decentralized digital
currencies. IACR Cryptology ePrint Archive 2015:464.
Vasek, M., Thornton, M. and Moore, T. 2014. Empirical
analysis of denial-of-service attacks in the bitcoin
ecosystem. In: Financial Cryptography and Data
Security, 8438:5771.
Wood, G. 2014. Ethereum: A secure decentralised
generalised transaction ledger. Yellow Paper, [online]
Available: http://gavwood.com/paper.pdf.
ICISSP 2018 - 4th International Conference on Information Systems Security and Privacy
138