Content uploaded by Adem Efe Gencer
Author content
All content in this area was uploaded by Adem Efe Gencer on Jan 17, 2018
Content may be subject to copyright.
Decentralization in Bitcoin and Ethereum
Networks
Adem Efe Gencer1,2,SoumyaBasu
1,2, Ittay Eyal1,3,RobbertvanRenesse
1,2,
and Emin Gün Sirer1,2
1Initiative for Cryptocurrencies and Contracts (IC3)
2Computer Science Department, Cornell University
3Electrical Engineering Department, Technion
Abstract. Blockchain-based cryptocurrencies have demonstrated how
to securely implement traditionally centralized systems, such as curren-
cies, in a decentralized fashion. However, there have been few measure-
ment studies on the level of decentralization they achieve in practice.
We present a measurement study on various decentralization metrics of
two of the leading cryptocurrencies with the largest market capitaliza-
tion and user base, Bitcoin and Ethereum. We investigate the extent of
decentralization by measuring the network resources of nodes and the
interconnection among them, the protocol requirements affecting the op-
eration of nodes, and the robustness of the two systems against attacks.
In particular, we adapted existing internet measurement techniques and
used the Falcon Relay Network as a novel measurement tool to obtain
our data. We discovered that neither Bitcoin nor Ethereum has strictly
better properties than the other. We also provide concrete suggestions
for improving both systems.
1Introduction
Cryptocurrencies are emerging as a new asset class, with a market capitalization
of about $150B as of Sept 2017 [15], a growing ecosystem, and a diverse commu-
nity. The most prominent platforms that account for over 70% of this market are
Bitcoin [57] and Ethereum [28, 70]. The underlying technology, the blockchain,
achieves consensus in a decentralized, open system and enables innovation in
industries that conventionally relied upon trusted authorities. Some examples of
such services include land record management [3], domain name registration [51],
and voting [55]. The key feature that empowers such services and makes these
platforms interesting is decentralization.Withoutit,suchservicesaretechnolog-
ically easy to construct but require trust in a centralized administrator.
Decentralization is a property regarding the fragmentation of control over the
protocol. In the Bitcoin and Ethereum protocols, users submit transactions for
miners to sequence into blocks. Better decentralization of miners means higher
resistance against censorship of individual transactions. For communication, Bit-
coin and Ethereum also have a peer-to-peer network for disseminating block and
transaction information. Both Bitcoin and Ethereum also contain full nodes,
arXiv:1801.03998v1 [cs.CR] 11 Jan 2018
which serve two critical roles: (1) to relay blocks and transactions to miners
(2) and to answer queries for end users about the state of the blockchain. Un-
derstanding the network properties of full nodes is crucial for protocol design and
analysis of each network’s resilience to attacks. Ongoing research explores ways
to make the Bitcoin and Ethereum networks more decentralized without mea-
surements on the underlying network. Hence, debates and decisions about the
underlying networks are often based on assumptions rather than measurement.
In this paper, we present a comprehensive measurement study on decentral-
ization metrics in these operational systems and shed light on whether or not
existing assumptions are satisfied in practice. We adapt prior Internet measure-
ment techniques for Bitcoin and Ethereum and use novel approaches to obtain
application layer data. Our main data sources are (1) direct measurements of
these networks from multiple vantage points, (2) aBitcoinrelaynetworkcalled
Falcon that we deployed and operated for a year, and (3) blockchain histories
of Bitcoin and Ethereum. Our study presents findings regarding the network
properties, impact of protocol requirements, security, and client interactions.
This paper makes three contributions. First, it provides new tools and tech-
niques for measuring blockchain-based cryptocurrency networks. The key tool
introduced here is the Falcon relay network that we built to serve as a backbone
for ferrying blocks. This network was deployed for Bitcoin across five continents,
providing a unique vantage point on pruned blocks. Second, we perform a com-
parative study of decentralization metrics in Bitcoin and Ethereum. Our key
findings are: (1) the Bitcoin network can increase the bandwidth requirements
for nodes by a factor of 1.7 and keep the same level of decentralization as 2016,
(2) the Bitcoin network is geographically more clustered than Ethereum, with
many nodes likely residing in datacenters. (3) Ethereum has lower mining power
utilization than Bitcoin and would benefit from a relay network, and (4) small
miners experience more volatility in block rewards in Bitcoin than Ethereum.
2BitcoinandEthereum
Bitcoin and Ethereum use Nakamoto consensus [5–7,57, 38] to regulate transac-
tion serialization in their blockchains. While architecturally very similar, these
systems differ significantly in terms of their API, abstractions, and wire protocol.
2.1 The Bitcoin Protocol
Bitcoin is a protocol that sequences transactions into groups called blocks. The
protocol targets a block production interval of 10 minutes with a maximum size
of 1 MB. At the time of our measurements, the last 100 blocks had a 0.99 MB
median block size and a 9.8 minute mean interval. The wire protocol implements
apeer-to-peernetworkbasedonfloodingblockandtransactionannouncements.
The peer to peer network is formed through point to point links. To form a
link, clients establish a TCP connection and perform a protocol-level three-way
handshake. The protocol-level handshake exchanges the state of each client, such
as the height of the blockchain and a version string associated with the software
Measurement Network Num. Nodes Dates
Bandwidth (All)
Latency (BTC IPv4)
(Single Beacon)
BTC IPv4 3441 Jan 11–16;Jan 30–Mar 16
IPv6 515 Jan 13–14;Apr 20–25
Tor 127 Jan 13;Apr 23–25
ETH IPv4 285 Mar 27–Apr 25
Peer-to-Peer Latency
(Mult. Vantage Pts.) BTC IPv4 3390 (5.7M edges) Jan 10–15; Jan 30–Mar 01
ETH IPv4 4302 (9.3M edges) Mar 01–Apr 11
Latency (Single Beacon) BTC IPv6 845 Jan 13–14; Feb 03–Apr 25
Pruned Blocks BTC IPv4 5977 May 5 2016 – Apr 29
Table 1: Timeline of measurements. All dates are in 2017 unless otherwise noted.
being run. When a client discovers or receives a new block, it floods the network
with the hash of the block. If a neighboring client needs the block, it requests the
block based on the hash value. There are many different block formats, such as
compact [17] and Merkle [44] blocks, but BMS focuses on retrieval of full blocks.
2.2 The Ethereum Protocol
The Ethereum protocol [28] focuses on providing a platform to facilitate building
decentralized applications on its blockchain. To sequence transactions, Ethereum
adopts a design inspired by Nakamoto consensus and the GHOST protocol [64].
Ethereum adopts a chain selection rule to harness the residual mining power
in pruned blocks for improved security. The protocol includes such blocks, called
uncles, in its blockchain and rewards the corresponding miners [70]. Ethereum
targets a block interval between 10 to 20 seconds [41]. The block size is indirectly
determined by an execution fee, called gas,thatfluctuatesovertime.Atthetime
of our measurements, the last 100 blocks were generated with a 2.9 KB median
block size and a 16.3 second average interval.
Fig. 1: The measurement infrastructure
is built on 18 globally distributed nodes.
Ethereum employs a UDP-based
node discovery mechanism inspired by
Kademlia [54], but the rest of the
P2P communication is over TCP.
Unlike Bitcoin, messages between
nodes are encrypted and authenti-
cated. Ethereum’s wire protocol is
poorly documented, so we rely on the
client implementations [42, 61, 19, 60]
and Ethereum wiki pages [26, 25, 29,
30, 27] for information.
In Ethereum, clients request blocks
by the corresponding block hash. Older clients request blocks, which consist of
a body and header, while newer clients request each piece separately. The mea-
surement system in this paper focuses on retrieval of full blocks and block bodies.
3MeasurementInfrastructure
Blockchain-based cryptocurrencies operate on global peer-to-peer networks that
span multiple administrative domains. Measurement of such networks concerns
the exploration of the relationship between peers, the capabilities of individual
peers, and the properties of the system as a whole–e.g. its security and fair-
ness. To characterize Bitcoin and Ethereum, we deployed Blockchain Measure-
ment System (BMS ), a measurement system than ran experiments of varying
duration–from a few days up to 12 months.
Network Properties. BMS uses multiple vantage points in order to gain a com-
prehensive view of the cryptocurrency networks. To capture the evolution of
these networks, BMS has been continuously collecting data regarding the pro-
visioned bandwidth of peers and peer-to-peer latency. BMS first connects to a
peer, collects measurements, and then disconnects before proceeding to the next
peer. These measurements target (1) Bitcoin nodes connected over IPv4, IPv6,
and Tor [23] and (2) Ethereum nodes connected over IPv4. As of May 2017,
Ethereum does not have any Tor nodes mainly because Tor is exclusively TCP,
whereas Ethereum node discovery is UDP-based. Moreover, this study excludes
Ethereum’s IPv6 network because BMS was unable to discover enough nodes to
reach generalized conclusions. Table 1 shows the timeline of the data collection
for each network and the number of nodes measured in each measurement.
To estimate the p eer-to-peer latency, BMS uses multiple vantage points geo-
graphically distributed across the world. Figure 1 shows the geographic distribu-
tion of the measurement infrastructure. 15 out of 18 nodes reside in PlanetLab’s
global research network [14] and the remaining three nodes are part of Cornell’s
academic network, located in Ithaca, NY.
To measure the provisioned bandwidth of nodes in Bitcoin and Ethereum,
BMS used nodes with extensive resources. In particular, measuring the maxi-
mum bandwidth that Bitcoin and Ethereum nodes have access to requires nodes
with (1) high download capacities to ensure that the bottlenecks are not in the
measurement apparatus, and (2) sufficient disk capacities to store detailed re-
sults. Since these machines need access to orders of magnitude higher bandwidth
capacity than what is achievable on shared infrastructure, such as PlanetLab
nodes, some BMS data was collected using dedicated, well-provisioned beacon
nodes located at Cornell University.
Finally, BMS needs to pick a sample of nodes from the Bitcoin and Ethereum
networks. As a sample, BMS uses a list containing nodes from Bitcoin and
Ethereum node crawling sites [1, 31], and a locally deployed Ethereum supernode
configured with a high peer limit. Interpretations in this paper assume that in-
ferences made from the reachable public nodes are representative of their entire
networks. In reality, these networks contains nodes that are not visible to the
public, e.g. they are behind a NAT or a firewall. One such class of nodes are part
of mining.Whilemuchofthemininginfrastructureisprivate,priormeasurement
work shows that mining operations often have gateway nodes to communicate
with the peer-to-peer network [56]. The properties of internal mining pool nodes
are orthogonal to the focus of this paper.
Blockchain Information. Anaiveapproachtoobtaininginformationaboutthe
blockchain would be to simply run a Bitcoin and Ethereum node. However,
this precludes information that cannot be obtained through the respective wire
protocols. Many important decentralization metrics center around the analysis
of blocks that are not part of the main blockchain. In Ethereum, many of these
blocks become uncles which can simply be requested through the wire protocol.
In Bitcoin, however, a block that is not part of the main blockchain simply
becomes pruned.PrunedblocksinBitcoinhavenoeffectonthestateofthe
system, they are deleted by clients without impacting correctness. Thus, it is
crucial to connect directly to miners to capture pruned blocks.
AcriticalcomponentofBMStoobserveprunedblocksistheFalconRelay
Network, which relays blocks between Bitcoin miners. The Falcon Relay Net-
work uses cut-through routing to quickly disseminate blocks worldwide, which
incentivizes miners to connect to Falcon. Indeed, Falcon is directly connected to
at least 36.4% of the entire hashpower in Bitcoin. Since there is just one other
operational relay network for Bitcoin [18,16], Falcon has observed blocks that
have not been seen on other well-connected nodes [8].
4Measurements
In this section, we present the measurements taken by BMS. In each measure-
ment, we describe the methodology, followed by the results of our analysis. As
with any measurement study of a large-scale, uninstrumentable artifact, mea-
surements are not perfect; we conclude each section by addressing some potential
sources of error and their mitigation.
4.1 Provisioned Bandwidth
Provisioned bandwidth is an estimate on a node’s transmission capacity charac-
terizing how much bandwidth the node has to communicate with the rest of the
cryptocurrency network. Greater provisioned bandwidth helps miners to propa-
gate/collect blocks to/from the network faster. Thus, it becomes more difficult
for a malicious miner to situate themselves in the network to achieve the rushing
property [35] and attack the blockchain. Knowledge of provisioned bandwidth
also aids in setting protocol parameters, such as the block size and frequency.
Methodology. BMS measures the provisioned bandwidth of each peer by re-
questing a large amount of data from each peer and seeing how fast the peers
can stream the data to BMS’s measurement nodes. BMS does this by asking for
blocks that were first seen over a year ago – similar to how a stale node asks
for blocks to sync state. Each request asks for the same set of blocks in Bitcoin
and blocks or the corresponding bodies in Ethereum. Next, BMS divides the
time into epochs and records the number of bytes received during each epoch.
This process continues until either BMS receives all data or a predefined timeout
of 30 seconds is reached. A long timeout helps BMS eliminate effects from TCP
slow start and other initialization noise as well as identify and eliminate spurious
spikes in throughput caused by buffering in the kernel by BMS. Finally, BMS
processes the collected data to determine the provisioned bandwidth. To do so,
first, it identifies the independent data streams by combining successive epochs
containing active data transfers. Then, it eliminates streams that are shorter
than 500 milliseconds to mitigate initialization artifacts such as TCP slow start.
BMS then outputs the maximum observed throughput among the remaining
distinct continuous streams as the provisioned bandwidth of the remote peer.
The experiments in this paper are run on servers with 1 Gbps links at Cornell
University. This has not changed from 2016 to 2017, which allows us to make
comparisons to a previous study in 2016 [20].
Bitcoin Eth.
IPv4 IPv6 Tor IPv4
[Mbps] [Mbps] [Mbps] [Mbps]
10% 5.7 11.0 2.1 3.4
33% 23.3 45.2 3.1 11.2
50% 56.1 78.2 4.1 29.4
67% 91.1 94.3 5.6 68.3
90% 177.0 207.9 8.1 144.4
Avg. 73.1 86.5 4.7 55.0
Std. Dev. 68.4 66.9 2.4 58.8
(a) Provisioned bandwidth statistics. (b) CDF
Fig. 2: Statistics on distribution of provisioned bandwidth and the CDF.
Results. Table 2a summarizes per-node bandwidth results that BMS has col-
lected. We see that Bitcoin nodes in both IPv4 and IPv6 networks have consis-
tently higher bandwidth compared to Ethereum IPv4 nodes. In particular, the
median Bitcoin IPv4 and IPv6 nodes have about 1.9⇥and 2.7⇥the bandwidth of
the typical Ethereum IPv4 node. In contrast, Bitcoin Tor nodes have an order of
magnitude lower bandwidth compared to directly connected nodes, though they
are not unusable – e.g. 90% has more than 2 Mbit/s. Ongoing research explores
alternatives to the Tor network that also provide efficient communication [50].
Figure 2b shows the cumulative distribution of the bandwidth measurements.
Steep increases in the Bitcoin IPv4/IPv6 curves at around 10 Mbps and 100
Mbps regions represent typical bandwidth capacities of a home user, and a typ-
ical Amazon EC2 Bitcoin instance. For Ethereum, we observe a similar accu-
mulation around 10 Mbps region, but the bandwidth is more evenly distributed
over the remaining nodes. As the long tailed distribution and higher standard
deviation indicates, bandwidth of Bitcoin IPv4/IPv6 nodes are spread out over
a wider range of values compared to Ethereum nodes. While the most well pro-
visioned Bitcoin nodes have around 300 Mbps of spare bandwidth, the highest
Ethereum node capacity that BMS has observed is limited to 250 Mbps.
One of the most interesting discoveries of this study is that the Bitcoin net-
work has improved tremendously in terms of its provisioned bandwidth. The re-
sults show that Bitcoin IPv4 nodes, which used to be connected to the network
with a median bandwidth of 33 Mbit/s in 2016 [20], have a median bandwidth
of 56 Mbit/s, as of February 2017. In other words, the provisioned bandwidth of
atypicalfullnodeisnow1.7⇥of what it was in 2016.
Critical system parameters, such as the maximum block size and block fre-
quency, can be increased when the provisioned bandwidth increases. The increase
in provisioned bandwidth suggests that the block size can be increased by a fac-
tor of 1.7 without increasing centralization beyond its de facto level in 2016.
Caveats. As with every measurement technique in the real world, our results
above are subject to experimental limitations and expected errors. The accu-
racy of the measurements may drop under certain circumstances, including the
cases where: (1) the network bottleneck lies on the side of the measurement
beacon rather than the remote peer, (2) network traffic on the side of BMS
interferes with the collected results, (3) the remote peer intentionally shapes
the traffic to selectively limit the bandwidth available to BMS, for instance via
bandwidth throttling, and (4) different steady state bandwidth between Bitcoin
and Ethereum, skewing the numbers for one system over another The setup of
our bandwidth infrastructure helps minimize potential inaccuracies due to the
first two issues. Moreover, analysis of popular Bitcoin [5] and Ethereum client
implementations [42,61, 19,60] shows that the third case is not supported by this
software and would require additional, potentially non-trivial, work to set up. To
verify the impact of the last issue, we ran an Ethereum and Bitcoin client and
saw that their bandwidth consumption differed by 0.2Mbps, which introduces
about a 1% error on our measurements above.
In addition to our analysis above, we also expect to see certain artifacts in our
data. As noted above, we see clusters of nodes around 10 Mbps and 100 Mbps,
which are typical bandwidth capacities of home and EC2 users, respectively.
4.2 Network Structure
The structure of the peer-to-peer network impacts the security and performance
for cryptocurrencies. A geographically clustered network can quickly propagate
a new block to many other nodes. This makes it more difficult for a malicious
miner to propagate conflicting blocks/transactions quicker than honest nodes.
However, a less clustered network may mean that full nodes are being run by a
wider variety of users which is also good for decentralization.
Methodology. Since it is not possible to obtain direct measurements between
peers we do not control, we use the state of the art estimation techniques to
establish bounds and gain insights into network structure.
Single Beacon Latency. We first collect direct ICMP ping measurements from
BMS nodes to all peers in the network. We report the minimum observed ping
latency, as it provides a physical bound on the distance to the BMS beacon.
Peer-to-Peer Latency. Measuring the peer-to-peer latency requires access to
the end points. In both Bitcoin and Ethereum, peers do not reveal their neigh-
bors. Hiding the network structure boosts privacy and security [45,56], but also
makes it harder to infer properties about the network. BMS provides latency
estimates for a superset of the actual links between known peers. We do not
normalize for the slightly different network sizes, 3390 for Bitcoin and 4302 for
Ethereum, as our samples from both networks were very similar. Since measuring
peer-to-peer latencies directly is not feasible, we establish bounds from observed
latencies from multiple beacons, using techniques from prior literature [37]. BMS
starts with the measurements taken from a single beacon. Then, it uses the tri-
angle inequality to estimate the upper and lower bounds for the latency between
peers. Repeating this process from other vantage points yields a set of bounds
for each pair of peers. Finally, BMS determines a range for latency estimates be-
tween each peer by picking the maximum lower bound and the minimum upper
bound. The paper also presents the average of the lower bound and upper bound
latency between peers. In this study, BMS includes nodes that do not support
the DAO fork [10] in its measurements for Ethereum.
Geographical Distance. BMS takes the minimum of repeated latency measure-
ments to eliminate transient network effects and capture the geographic distance
between two nodes [43, 13, 69]. BMS also uses IP geolocation data to calculate
distances between peer nodes as an additional validation on our results. To cal-
culate distances, BMS applies the Haversine formula [63] using the coordinate
values gathered from an IP-based geolocation service [46].
Single Beacon Peer-to-Peer
Bitcoin Bitcoin Eth.
IPv4 IPv6 IPv4 IPv4
[ms] [ms] [ms] [ms]
10% 29 40 48 92
33% 78 80 79 125
50% 89 95 109 152
67% 98 95 152 200
90% 201 165 286 276
Avg. 97 103 135 171
Std. Dev. 59 62 88 76
Table 2: Min single b eacon latencies o b-
served and P2P latency estimates.
Results. Our measurements indicate
considerable differences between P2P
latencies of Bitcoin and Ethereum
IPv4 networks, summarized in Table 2
and PDF graphed in Figure 3.
We find that Bitcoin ha s ma ny
more nodes that are closer geograph-
ically than Ethereum. Figure 3 shows
that Ethereum’s most likely laten-
cies are centered around 120ms, while
Bitcoin nodes tend to be clustered
around 50ms. Only 13% of Ethereum
latencies are under 100ms, while Bit-
coin has a surprisingly high 46%. Ad-
ditionally, the estimated peer-to-peer
latency between Ethereum nodes is 26.7% higher than Bitcoin on average. This
geographic proximity between nodes, along with the observation that Bitcoin has
many nodes with 100 Mbps of provisioned bandwidth (see Section 4.1), seems to
indicate that many Bitcoin nodes are run in datacenters. 56% of Bitcoin’s nodes
and 28% of Ethereum’s nodes belong to an autonomous system that provides
dedicated hosting services, a difference significant at the 1% significance level.
Indeed Ethereum nodes are not accumulated in a single geographical region,
but are more evenly distributed around the world. Figure 3c shows the CDF of
distances between peer to peer nodes based on IP geolocation information. The
results corroborate our findings based on network latency measurements and
show that Ethereum nodes are geographically further apart than Bitcoin. As
additional evidence, when we use geolocation on the P2P distances and plot the
CDF in Figure 3c, we see that Ethereum nodes are further apart than Bitcoin.
(a) (b) (c)
Fig. 3: The histogram of P2P latencies in Bitcoin (Fig 3a) and Ethereum (Fig 3b),
as well as the CDF of geographical distances (Fig 3c).
Sanity Checks. The first two columns of Table 2 present single beacon latency
in Bitcoin IPv4/IPv6 networks. The results indicate that both the median and
the average latency to IPv4 nodes are smaller than IPv6 nodes. As there are
fewer IPv6 nodes than IPv4 nodes, we expect this result since IPv4 nodes are
more likely to be closer to our beacons.
While there has been a large body of work showing the prevalence of triangle
inequality violations in the Internet [52,67, 12], there are several reasons BMS’s
measurements are not affected significantly. First, such violations were shown
to occur less than 5% of network snapshots [52]. Since we take the minimum
latency observed from a beacon, triangle inequality violations will only occur in
our dataset less than 1% of the time [52]. TIVs are also significantly less prevalent
when dealing with latencies less than 300 ms, which includes almost our entire
dataset [67]. To ensure that the above results hold for our dataset as well, we
used a geolocation service as ground truth to verify our results.
One other limitation in our study is that it is impossible to collect measure-
ments using ICMP pings from nodes that block ICMP traffic and from Tor nodes
that only communicate over TCP.
4.3 Distribution of Mining Power
Mining on cryptocurrency networks is a complex process that typically requires
large computation power. With the current mining difficulty of Bitcoin and
Ethereum, using commodity hardware to generate blocks is not feasible [21]
which centralizes the mining process somewhat. However, as long as there are
many different entities mining, the system is still decentralized. We compare the
decentralization of the mining process between Bitcoin and Ethereum.
Methodology. To identify the power of miners in Bitcoin and Ethereum, we
examined their weekly distribution over the last 10 months starting on July 15,
2016. Our mining power estimations are based on the ratio of main chain blocks
generated by distinct entities. Hence, pruned blocks in Bitcoin and uncles in
Ethereum do not affect these estimations. In both networks, miners voluntarily
disclose their identity as part of each block they mine. We gathered this data
from a public API for Bitcoin [9] and a blockchain explorer for Ethereum [32].
In Bitcoin, 1.8% of the blocks were unidentified, which we treated as if they
were generated by distinct individual miners. Finally, we manually processed
identities to detect and merge duplicates. This includes pools operated by the
same administrator [47] and multiple identities representing the same pool, which
5
15
25 Ethereum
Ratio of Mining Power (%)
5
15
25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Bitcoin
Miner Index (Descending Mining Power)
Fig. 4: Distribution of mining power in Bitcoin and Ethereum networks. Bars
indicate observed standard deviation from the average.
we identified by looking for the same pool name with a corresponding tag, e.g.
’DwarfPool1’ and ’DwarfPool2’. We do not distinguish between solo miners and
mining pools, because it would be misleading to do so. Some claim that mining
pools provide a modicum of decentralization, albeit invisible, because they are
comprised of multiple solo miners who can exit the pool if the pool operator
misbehaves. While this is true, it requires detection of misbehaviors, which is
not always possible. But in any case, our study captures a historical account
of the mining power distribution as it took place, and at the time a block was
committed to the chain, the pool constituents were plaintively cooperating as
part of the same pool.
Results. For each week of the analys is p eriod, we calculated the corresp onding
mining power of entities and ranked each miner accordingly. Figure 4 shows the
top 20 weekly mining power distribution in the Ethereum and Bitcoin networks.
Each group of bars represents a chronologically ordered collection of weekly
mining power ratios,definedasthefractionofblockscontributedbyaminer.
Figure 4 illustrates that, in Bitcoin, the weekly mining power of a single entity
has never exceeded 21% of the overall power. In contrast, the top Ethereum miner
has never had less than 21% of the mining power. Moreover, the top four Bitcoin
miners have more than 53% of the average mining power. On average, 61% of
the weekly power was shared by only three Ethereum miners. These observations
suggest a slightly more centralized mining process in Ethereum.
Although miners do change ranks over the observation period, each spot
is only contested by a few miners. In particular, only two Bitcoin and three
Ethereum miners ever held the top rank. The same mining pool has been at
the top rank for 29% of the time in Bitcoin and 14% of the time in Ethereum.
Over 50% of the mining power has exclusively been shared by eight miners in
Bitcoin and five miners in Ethereum throughout the observed period. Even 90%
of the mining power seems to be controlled by only 16 miners in Bitcoin and
only 11 miners in Ethereum. Hence, both platforms rely heavily on very few
distinct mining entities to maintain the blockchain. Indeed, we see in Figure 5
that the mining power trends can be fit as exponential distributions with curves
0.21e0.19xand 0.35e0.30xin Bitcoin and Ethereum, respectively. These curves
yield a coefficient of determination value of 0.99.
These results show that a Byzantine quorum system [53] of size 20 could
achieve better decentralization than proof-of-work mining at a much lower re-
source cost. This shows that further research is necessary to create a permission-
less consensus protocol without such a high degree of centralization.
5
15
25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Ratio of
Mining Power (%)
Miner Index (Descending Mining Power)
Ethereum
Bitcoin
Fig. 5: Exponential trendlines for the av-
erage distribution of mining power.
Sanity Checks. Similar to other
works in the literature [58, 68], we
assume that miners accurately self-
identify themselves. A miner that con-
tributes a significant portion of the
hash power to the cryptocurrency can
exert some amount of influence over
protocol changes. Thus, it is likely
that miners will want to claim blocks
that they generated as their own.
While strong miners gain political clout and attract more members, getting
too large raises alarms among the community about centralization. Thus, such
miners may conceal or obfuscate this information to appear less powerful – e.g.
by generating multiple identities. For instance, two major mining pools, Ethpool
and Ethermine, publicly reveal that they share the same admin [47]. Thus, any
analysis based on the voluntary miner data skews toward a more decentralized
network than the reality.
4.4 Mining Power Utilization
Mining power utilization [34], which measures the fraction of mined blocks that
remain in the main chain, is a metric for evaluating the efficiency of a protocol, as
well as a second order metric for robustness against rollbacks. As mining power
utilization increases, the protocol is able to convert more of the energy spent to
useful work, and therefore the cost to launch an attack is higher.
Methodology. To study the mining power util ization, we analyzed weekly a nd
daily distribution of pruned blocks in Bitcoin and uncles in Ethereum, compared
to the main chain blocks. While uncle blocks in Ethereum contribute to the
selection of the main chain and reward miners, uncle blocks do not sequence
transactions. Thus, uncle blocks do not contribute to Ethereum’s performance,
but do contribute to the blockchain’s security. We retrieved this data from (1) the
Falcon networ k, (2) alocalBitcoinclient,and(3) public blockchain explorers
for Bitcoin [9] and Ethereum [32]. In particular, the Bitcoin blockchain explorer
and Falcon exclusively provided 12% and 20% of the total 124 pruned blocks,
respectively. Both sources commonly discovered the remaining 68%.
Results. Figure 6a and Figure 6b show weekly and daily distributions of min-
ing power utilization in Bitcoin and Ethereum networks, respectively. The re-
sults show that Bitcoin utilization is always above 99%, which means that a
pruned block in Bitcoin is a relatively rare event. In contrast, daily utilization in
Ethereum is typically between 90% to 94% range and never goes above the 97%
(a) Bitcoin- weekly MPU from 99% (b) Ethereum- daily MPU from 70%
Fig. 6: Mining power utilization (MPU) for Bitcoin and Ethereum
threshold. During 2016, Ethereum faces occasional drops in its utilization down
from 74% to 88%, including (1) the days following the exploitation of the DAO
vulnerability [10] from June 17 to 18, (2) attacks on Ethereum network [11,66]
between September 22 to October 19, and (3) the days following the Spurious
Dragon hard fork [48] between November 23 to 29. These results indicate a strong
relationship between mining power utilization and real life events in Ethereum.
This may be the result of preventive measures that spam the network to slow
down the DAO attacker, bad actors generating blocks with excessive resource
demands, and miners with outdated clients. These results indicate that a relay
network, like Falcon, would be greatly beneficial to the Ethereum network.
Sanity Checks. The design of the Ethereum protocol requires peers to store and
propagate uncle blocks, which are not on the main chain. In contrast, Bitcoin’s
blockchain only stores the main chain so peers do not propagate pruned blocks.
Hence, capturing such blocks in Bitcoin requires actively watching the network.
While the Falcon relay network provides a strong incentive to miners to relay
blocks through it, some miners may choose not to do so. Consequently, we may
be missing some pruned blocks that were generated by the Bitcoin network.
4.5 Fairness
1
4
7Ethereum
Fairness Ratio
(log scale)
1
4
7
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Bitcoin
Miner Index (Descending Mining Power)
Fig. 7: Fairness distribution. Zero fair-
ness means no pruned block from miner.
Section 4.3 presented the mining
power distribution, which looks at
the main chain presence of miners.
The impact of this distribution on a
miner’s pruned block rate is unclear.
To study this relationship, we exam-
ine fairness defined as the ratio of a
miner’s share of pruned blocks to her
mining power. In a fair protocol, min-
ers generate pruned blocks proportional to their mining power; hence, the fairness
is close to 1. A fairness greater than 1 implies that the miner is at a disadvantage,
while a fairness less than 1 implies that the miner has an advantage.
Methodology. We used the Falcon network, and a Bitcoin blockchain ex-
plorer [8] to retrieve pruned Bitcoin blocks. These sources have, respectively,
provided 109 and 99 blocks, yielding a total of 124 distinct pruned blocks. We
collected uncles from an Ethereum blockchain explorer [32].
Similar to Section 4.3, our results here also assume that miners voluntarily
identify themselves in uncles/pruned blocks. Another caveat here lies in gather-
ing pruned blocks. While we incentivize miners to relay blocks through Falcon,
there is no guarantee that they necessarily will do so. We suspect that explicit
storage of uncles in Ethereum captures a larger proportion of pruned blocks.
Results. Figure 7 shows the distribution of fairness of 20 miners with the highest
mining power. The results indicate that, in both networks, the top four miners
generally are more successful at appending blocks to the main chain. We run
the Kolmogorov-Smirnov goodness of fit test with a p-value of 0.01 to compare
the fairness distributions of Bitcoin and Ethereum. Perhaps surprisingly, we see
that the fairness of Ethereum and Bitcoin differ significantly from each other
keeping a constant time period. The reason for this difference is a much larger
standard deviation in Bitcoin’s miner fairness compared to Ethereum (1.72 ver-
sus 0.25). The mean of both fairness distributions, however, are very similar,
with Ethereum at 1.08 and Bitcoin at 1.22.
A high variance results in centralization pressure since smaller miners will
have a more difficult time affording the loss of revenue due to a transiently
high fairness score. This high variance is a direct result of a significantly smaller
number of blocks being generated in Bitcoin. Since Ethereum has a higher block
frequency than Bitcoin, smaller miners have a more predictable payoffthan larger
miners. This makes Ethereum more predictable to mine for smaller miners due
to the lower variance in block rewards. Thus, it is important for blockchain
protocols to take variance of the block rewards in addition to the mean.
Simply increasing the block frequency may not be the solution to decrease the
variance of block rewards since the mining power distribution may be affected
as well. The increased block frequency in Ethereum may be part of the cause of
the slightly more centralized mining power distribution (see Section 4.3).
Sanity Checks. Similar to Section 4.4, our results here also assume that
miners voluntarily identify themselves in uncles/pruned blocks. As before, if
the miners are lying, they are likely to present a more fair system than reality.
Another caveat here lies in gathering pruned blocks. While we incentivize miners
to relay blocks through Falcon, there is no guarantee that they will. We suspect
that explicit storage of uncles in Ethereum allows for more accurate analysis.
Finally, Bitcoin has a significantly lower block generation frequency than
Ethereum. On top of that, Bitcoin also has a lower pruned block rate than
Ethereum does, which means it has significantly fewer pruned blocks. Thus, this
fairness metric is much noisier in Bitcoin compared to Ethereum.
5RelatedWork
Network measurements in blockchain-based systems have mainly focused on Bit-
coin. One such study [22] demonstrated that the latency is the dominating factor
in propagation of blocks smaller than 20 KB. Following work [20] has shown that
(1) this limit has increased to 80 KB and (2) nodes are provisioned with sub-
stantially higher bandwidth capacity than what the protocol demands. Feld et
al. [36] pointed out a strong AS-level centralization that may impact Bitcoin
network’s connectivity – i.e. 10 ASes contain over 30% of peers. Recent work [2]
presented the level of vulnerability, showing that 13 ASes cover the same fraction
of peers, but only 39 IP prefixes host half of the overall mining power. Ours is
the first work that does a similar type of study on Ethereum as well.
Other work studied various aspects of the Bitcoin overlay network. Miller et
al. [56] found that a small fraction of the network, containing around 100 nodes,
represents more than 75% of the mining power. The study conjectured that these
nodes are well-connected to major mining pools; hence, provide higher efficiency
in broadcasting blocks. Biryukov et al. [4] examined how peer neighbors discover
IP addresses that correspond to pseudonymous identities. Another study [49]
deanonymized peers by observing anomalous relaying behavior in network. Pap-
palardo et al. [59] observed that low value transactions may experience waiting
times of over a month. Other work measured churn and geolocated peers [24].
Gervais et al. [40] discussed centralization concerns regarding the client develop-
ment process, distribution of mining power, and spendable coins. Most of these
works focus on attacks and the structure of the overlay network, while this work
focuses on the resource capabilities of the nodes used in the overlay network.
Recent work presented ways to reduce resource requirements to participate in
blockchain systems. Such solutions enhance decentralization by increasing the di-
versity of participants. Aspen [39] achieves this through sharding the blockchain.
In this system, users store, process, and propagate only the data that is relevant
to them, hence need fewer resources to join the network. Another approach [62]
relies on authenticated data structures to reduce load on nodes. Relay net-
works increase network efficiency through faster block propagation. The first
such system [16] achieved this by avoiding full block verification and retransmit-
ting known transactions. Falcon, the source of pruned block data in the Bitcoin
network in this paper, relies on cut-through routing for faster block propagation.
Finally, FIBRE incorporates cut-through routing with compact blocks [17] and
forward error correction over UDP. The novelty in our work was utilizing Falcon
data in order to gain insights into transient application layer information.
Blockchain explorers [65,8, 32,33] provide a variety of data on cryptocurrency
networks, including online blockchain history; statistics on blockchain compo-
nents, transaction fees, and market value; and node information. While these
sources of information are useful to the community, this work scientifically tests
whether the intuitions provided by these sources of information indeed hold.
6Conclusion
Decentralization in blockchain-based platforms is a component of the value
proposition these systems offer. This work presents a comparative assessment of
decentralization in two most popular cryptocurrencies, Bitcoin and Ethereum.
To do so, it relies on novel measurement techniques to obt ain application layer
information using the Falcon Network and the application of well-established
internet measurement techniques.
Our observations show that Bitcoin has a higher capacity network than
Ethereum,but with more clustered nodes likely in datacenters. We also observe
that Bitcoin and Ethereum have fairly centralized mining processes and that
further research is needed to decentralize permissionless consensus protocols fur-
ther. In Ethereum, the block rewards have less variance than Bitcoin’s. Finally,
Ethereum has a lower mining power utilization than Bitcoin, likely due to the
high block frequency.
Further, we see that Bitcoin has undergone tremendous growth and can in-
crease the block size by a factor of 1.7x without any decrease in decentralization
compared to 2016. Additionally, our study uncovers that the volatility of mining
rewards is an important, but often ignored, metric. Finally, we see that Ethereum
would likely benefit from a relay network to increase its mining power utilization.
7Acknowledgements
The authors thank Vitalik Buterin and the anonymous reviewers for their feed-
back on earlier drafts of this manuscript. Ittay Eyal is supported by the Viterbi
Fellowship in the Center for Computer Engineering at the Technion. This work
was partially funded and supported by AFOSR grant F9550-16-0250, NSF CSR-
1422544, NSF CNS-1601879, NSF CNS-1544613, NSF CCF-1522054, NSF CNS-
1518779, NSF CNS-1704615, ONR N00014-16-1-2726, NIST Information Tech-
nology Laboratory (60NANB15D327, 70NANB17H181), Facebook, Infosys, and
IC3, the Initiative for Cryptocurrencies and Smart Contracts. This material is
based upon work supported by the National Science Foundation Graduate Re-
search Fellowship Program under Grant No. DGE-1650441. Any opinions, find-
ings, and conclusions or recommendations expressed in this material are those
of the authors and do not necessarily reflect the views of the National Science
Foundation.
References
1. 21.CO. Bitnodes. https://bitnodes.21.co/, retrieved Jun. 2017.
2. M. Apostolaki, A. Zohar, and L. Vanbever. Hijacking Bitcoin: Routing attacks on
cryptocurrencies. arXiv preprint arXiv:1605.07524,2016.
3. Benben Team. Benben. http://benben.com.gh/,retrievedOct.2016.
4. A. Biryukov, D. Khovratovich, and I. Pustogarov. Deanonymisation of clients in
Bitcoin P2P network. In Proc. of the ACM Conference on Computer and Commu-
nications Security (CCS),pages15–29,2014.
5. Bitcoin Community. Bitcoin source. https://github.com/bitcoin/bitcoin,re-
trieved Jun. 2017.
6. Bitcoin Community. Protocol rules. https://en.bitcoin.it/wiki/Protocol_
rules, retrieved Jun. 2017.
7. Bitcoin Community. Protocol specification. https://en.bitcoin.it/wiki/
Protocol_specification, retrieved Jun. 2017.
8. Blockchain Info Team. Blockchain Info. https://blockchain.info/,retrieved
May. 2017.
9. BlockTrail Team. Blocktrail API. https://www.blocktrail.com/api/docs#api_
data,retrievedApr.2017.
10. V. Buterin. Critical update re: DAO vulnerability. https://blog.ethereum.org/
2016/06/17/critical-update-re-dao-vulnerability/,retrievedApr.2017.
11. V. Buterin. Transaction spam attack: Next steps. https://blog.ethereum.org/
2016/09/22/transaction-spam-attack-next-steps/,retrievedApr.2017.
12. F. Cangialosi, D. Levin, and N. Spring. Ting: Measuring and exploiting latencies
between all Tor nodes. In Proc. of the ACM SIGCOMM Conference on Internet
Measurement (IMC),pages289–302,2015.
13. B. Chandrasekaran, M. Bai, M. Schoenfield, A. Berger, N. Caruso, G. Economou,
S. Gilliss, B. Maggs, K. Moses, D. Duff, et al. Alidade: Ip geolocation without
active probing. Technical report, Duke University, 2015.
14. B. Chun, D. Culler, T. Roscoe, A. Bavier, L. Peterson, M. Wawrzoniak, and
M. Bowman. PlanetLab: an overlay testbed for broad-coverage services. ACM
SIGCOMM CCR,33(3):3–12,2003.
15. CoinMarketCap. Cryptocurrency market capitalizations. https:
//coinmarketcap.com/,retrievedMay.2017.
16. M. Corallo. The Bitcoin relay network. BIP 152, http://bitcoinrelaynetwork.
org/,retrievedMay.2017.
17. M. Corallo. Compact block relay. BIP 152, https://github.com/bitcoin/bips/
blob/master/bip-0152.mediawiki,retrievedJun.2017.
18. M. Corallo. FIBRE: Fast internet Bitcoin relay engine. https://github.com/
bitcoinfibre/bitcoinfibre,retrievedApr.2017.
19. Cpp-ethereum Authors. Ethereum C++ client. https://github.com/ethereum/
cpp-ethereum,retrievedApr.2017.
20. K. Croman, C. Decker, I. Eyal, A. E. Gencer, A. Juels, A. Kosba, A. Miller, P. Sax-
ena, E. Shi, E. G. Sirer, D. Song, and R. Wattenhofer. On scaling decentralized
blockchains (a position paper). In Proc. of the Workshop on Bitcoin and Blockchain
Research (BITCOIN),Barbados,2016.
21. Cryptocompare Team. Cryptocurrency mining calculator. https://www.
cryptocompare.com/mining/calculator, retrieved Jun. 2017.
22. C. Decker and R. Wattenhofer. Information propagation in the Bitcoin network.
In Proc. of the IEEE International Conference on Peer-to-Peer Computing,pages
1–10, Trento, Italy, 2013.
23. R. Dingledine, N. Mathewson, and P. Syverson. Tor: The second-generation onion
router. In Proc. of the USENIX Security Symposium,2004.
24. J. A. D. Donet, C. Pérez-Sola, and J. Herrera-Joancomartí. The Bitcoin P2P
network. In Proc. of the International Financial Cryptography and Data Security
Conference,pages87–102,Barbados,2014.
25. Ethereum Community. devp2p forward compatibility requirements for home-
stead. https://github.com/ethereum/EIPs/blob/master/EIPS/eip-8.md,re-
trieved Apr. 2017.
26. Ethereum Community. D⌅Vp2p wire protocol. https://github.com/ethereum/
wiki/wiki/%C3%90%CE%9EVp2p-Wire-Protocol,retrievedApr.2017.
27. Ethereum Community. Ethereum wire protocol. https://github.com/ethereum/
wiki/wiki/Ethereum-Wire-Protocol,retrievedApr.2017.
28. Ethereum Community. A next generation smart contract and decentralized ap-
plication platform. https://github.com/ethereum/wiki/wiki/White-Paper,re-
trieved Apr. 2017.
29. Ethereum Community. RLPx: Cryptographic network & transport proto-
col. https://github.com/ethereum/devp2p/blob/master/rlpx.md,retrieved
Apr. 2017.
30. Ethereum Community. RLPx encryption. https://github.com/ethereum/go-
ethereum/wiki/RLPx-Encryption,retrievedApr.2017.
31. EthereumJ. The Ethereum nodes explorer. https://www.ethernodes.org/,re-
trieved Jun. 2017.
32. Etherscan Team. Etherscan: The Ethereum block explorer. https://etherscan.
io/, retrieved Jun. 2017.
33. Ethstats Team. Ethstats. https://ethstats.net/, retrieved Jun. 2017.
34. I. Eyal, A. E. Gencer, E. G. Sirer, and R. van Renesse. Bitcoin-NG: A scalable
blockchain protocol. In Proc. of the USENIX Symposium on Networked Systems
Design and Implementation (NSDI),pages45–59,SantaClara,CA,USA,2016.
35. I. Eyal and E. G. Sirer. Majority is not enough: Bitcoin mining is vulnerable. In
Proc. of the International Financial Cryptography and Data Security Conference,
Barbados, 2014.
36. S. Feld, M. Schönfeld, and M. Werner. Analyzing the deployment of Bitcoin’s P2P
network under an AS-level perspective. Proc. of the International Workshop on
Secure Peer-to-Peer Intelligent Networks and Systems,32:1121–1126,2014.
37. P. Francis, S. Jamin, C. Jin, Y. Jin, D. Raz, Y. Shavitt, and L. Zhang. Idmaps:
Aglobalinternethostdistanceestimationservice. IEEE/ACM Transactions on
Networking (TON),pages525–540,2001.
38. J. A. Garay, A. Kiayias, and N. Leonardos. The Bitcoin backbone protocol: Anal-
ysis and applications. In Proc. of the International Conference on the Theory
and Applications of Cryptographic Techniques (EUROCRYPT),pages281–310.
Springer, 2015.
39. A. E. Gencer, R. van Renesse, and E. G. Sirer. Short paper: Service-oriented
sharding for blockchains. In Proc. of the International Financial Cryptography and
Data Security Conference,Sliema,Malta,2017.
40. A. Gervais, G. O. Karame, V. Capkun, and S. Capkun. Is Bitcoin a decentralized
currency? Proc. of the IEEE Symposium on Security and Privacy,3(12):54–60,
2014.
41. A. Gervais, G. O. Karame, K. Wüst, V. Glykantzis, H. Ritzdorf, and S. Capkun.
On the security and performance of proof of work blockchains. In Proc. of the
ACM Conference on Computer and Communications Security (CCS),pages3–16,
Vienna, Austria, 2016.
42. Go-ethereum Authors. Official Go implementation of the Ethereum protocol.
https://github.com/ethereum/go-ethereum,retrievedApr.2017.
43. S. Guha, R. Murty, and E. G. Sirer. Sextant: A unified framework for node and
event localization in sensor networks. In Proc. of the ACM International Sympo-
sium on Mobile Ad Hoc Networking and Computing,pages205–216.ACM,2005.
44. M. Hearn and M. Corallo. Connection bloom filtering. BIP 37, https://github.
com/bitcoin/bips/blob/master/bip-0037.mediawiki,retreivedSept.2017.
45. E. Heilman, A. Kendler, A. Zohar, and S. Goldberg. Eclipse attacks on Bitcoin’s
peer-to-peer network. In Proc. of the USENIX Security Symposium,pages129–144,
Washington, D.C., USA, 2015.
46. IP Info Team. IP Info. http://ipinfo.io/,retrievedApr.2017.
47. jackwinters. Ethpool & Ethermine voting on the soft fork. https://
forum.daohub.org/t/ethpool-ethermine-voting-on-the-soft-fork/5364,re-
trieved Apr. 2017.
48. H. Jameson. Hard fork no. 4: Spurious Dragon. https://blog.ethereum.org/
2016/11/18/hard-fork-no-4-spurious-dragon/,retrievedApr.2017.
49. P. Koshy, D. Koshy, and P. McDaniel. An analysis of anonymity in Bitcoin using
P2P network traffic. In Proc. of the International Financial Cryptography and
Data Security Conference,pages469–485,Barbados,2014.
50. A. Kwon, D. Lazar, S. Devadas, and B. Ford. Riffle: An efficient communica-
tion system with strong anonymity. Proc. of the Privacy Enhancing Technologies
Symposium (PETS),2016.
51. A. Loibl. Namecoin. namecoin.info,2014.
52. C. Lumezanu, R. Baden, N. Spring, and B. Bhattacharjee. Triangle inequality
variations in the internet. In Proc. of the ACM SIGCOMM Conference on Internet
Measurement (IMC),pages177–183,2009.
53. D. Malkhi and M. Reiter. Byzantine quorum systems. Journal of Distributed
Computing,11(4):203–213,1998.
54. P. Maymounkov and D. Mazieres. Kademlia: A peer-to-peer information system
based on the XOR metric. In Proc. of the International Workshop on Peer-to-Peer
Systems,pages53–65.Springer,2002.
55. P. McCorry, S. Shahandashti, and F. Hao. A smart contract for boardroom voting
with maximum voter privacy. In Proc. of the International Financial Cryptography
and Data Security Conference,Sliema,Malta,2017.
56. A. Miller, J. Litton, A. Pachulski, N. Gupta, D. Levin, N. Spring, and B. Bhat-
tacharjee. Discovering Bitcoin’s public topology and influential nodes, 2015.
57. S. Nakamoto. Bitcoin: A peer-to-peer electronic cash system, 2008.
58. K. Nayak, S. Kumar, A. Miller, and E. Shi. Stubborn mining: Generalizing selfish
mining and combining with an eclipse attack. IACR Cryptology ePrint Archive,
2015:796, 2015.
59. G. Pappalardo, T. Di Matteo, G. Caldarelli, and T. Aste. Blockchain inefficiency
in the Bitcoin peers network. arXiv preprint arXiv:1704.01414,2017.
60. Parity Authors. Ethereum Rust client. https://github.com/paritytech/parity,
retrieved Apr. 2017.
61. Pyethapp Authors. Python based client implementing the Ethereum protocol.
https://github.com/ethereum/pyethapp/,retrievedApr.2017.
62. L. Reyzin, D. Meshkov, A. Chepurnoy, and S. Ivanov. Improving authenticated
dynamic dictionaries, with applications to cryptocurrencies. In Proc. of the In-
ternational Financial Cryptography and Data Security Conference,Sliema,Malta,
2017.
63. R. W. Sinnott. Virtues of the haversine. Sky and Telescope,1984.
64. Y. Sompolinsky and A. Zohar. Secure high-rate transaction processing in Bitcoin.
In Proc. of the International Financial Cryptography and Data Security Conference,
pages 507–527, Puerto Rico, 2015.
65. Statoshi Team. Statoshi info. http://statoshi.info/,retrievedMay.2017.
66. M. Swende. Announcement of imminent hard fork for EIP150 gas cost
changes. https://blog.ethereum.org/2016/10/13/announcement-imminent-
hard-fork-eip150-gas-cost-changes/,retrievedApr.2017.
67. G. Wang, B. Zhang, and T. Ng. Towards network triangle inequality violation
aware distributed systems. In Proc. of the ACM SIGCOMM Conference on Internet
Measurement (IMC),pages175–188.ACM,2007.
68. L. Wang and Y. Liu. Exploring miner evolution in Bitcoin network. In Proc. of
the International Workshop on Passive and Active Network Measurement,pages
290–302. Springer, 2015.
69. Y. Wang, D. Burgener, M. Flores, A. Kuzmanovic, and C. Huang. Towards street-
level client-independent ip geolocation. In Proc. of the USENIX Symposium on
Networked Systems Design and Implementation (NSDI),pages365–379.USENIX
Association, 2011.
70. G. Wood. Ethereum: A secure decentralised generalised transaction ledger.
Ethereum Project Yellow Paper,2014.
