Papers
Topics
Authors
Recent
Search
2000 character limit reached

Decentralization in Bitcoin and Ethereum Networks

Published 11 Jan 2018 in cs.CR | (1801.03998v2)

Abstract: Blockchain-based cryptocurrencies have demonstrated how to securely implement traditionally centralized systems, such as currencies, in a decentralized fashion. However, there have been few measurement 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 capitalization 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 operation 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.

Citations (375)

Summary

  • The paper reveals that both Bitcoin and Ethereum face significant centralization in mining and hosting, with Bitcoin showing slightly better network dispersion.
  • The study employs longitudinal data, node discovery, and packet analysis to quantify metrics like bandwidth, latency, and orphan block rates.
  • Findings indicate inherent trade-offs between economic incentives and decentralization, suggesting a need for advanced protocol design and improved network engineering.

Decentralization Analysis of Bitcoin and Ethereum Networks

Introduction

"Decentralization in Bitcoin and Ethereum Networks" (1801.03998) offers a comprehensive empirical investigation into the structure and operational features of Bitcoin and Ethereum, evaluating the degree and nature of decentralization within these two major permissionless blockchain systems. The study quantitatively examines networking, mining, and protocol-level characteristics, employing longitudinal data collection from geographically distributed nodes, packet analysis, and mining telemetry.

Methodology

The authors adopt multiple measurement heuristics. Network-level decentralization is assessed through node discovery, latency measurement, bandwidth analysis, and peer connection topology inference. Mining decentralization is investigated using block propagation timing, miner block shares, empty block rates, and orphaned block analysis. The data encompasses months of active monitoring via custom distributed measurement infrastructures, providing statistically significant insight with respect to sampling bias and measurement precision.

Key Findings

Network Layer Properties

The paper demonstrates that both Bitcoin and Ethereum display relatively high node churn rates and varying degrees of bandwidth provisioning. Measured Bitcoin nodes generally exhibit higher bandwidth and lower latency than Ethereum peers, suggesting more robust global network infrastructure. Peer connection graph analyses indicate a propensity for clustering, partially undermining strict egalitarian network topologies. Geographical distribution metrics further reveal substantial centralization in major hosting providers and data centers, with Bitcoin possessing generally greater dispersion than Ethereum.

Mining and Block Propagation

Both systems rely on a small set of dominant mining entities: the top four miners in Bitcoin and Ethereum together control a majority of the computational resources and block production, substantiating concerns about mining centralization. Block propagation measurements show Bitcoin benefits from compact block relay and efficient orphans suppression, yielding reduced propagation delay and lower orphan rates compared with Ethereum, where propagation times and orphan rates are consistently elevated due to protocol design choices.

Protocol-Level Decentralization

The authors find that economic incentives (e.g., block rewards, uncle inclusion) profoundly affect decentralization outcomes. Ethereum’s incentivization for including uncle blocks both mitigates and distorts its network-level orphaning characteristics. The analysis asserts that both systems must contend with trade-offs between network decentralization, mining concentration, and transaction propagation performance, with neither network fulfilling all decentralization ideals simultaneously.

Numerical Results

Quantitative results are especially robust:

  • In Bitcoin, median node bandwidth exceeds 50 Mbps, with block propagation latencies below 500 ms for well-connected peers.
  • For Ethereum, median bandwidth is ~10 Mbps, and propagation latencies are higher (often >1 s).
  • The top four mining pools in Bitcoin produce over 53% of blocks; in Ethereum, the top four exceed 61%.
  • Orphaned block rate in Ethereum is approximately 6%, while Bitcoin’s orphan/block rate remains near 0.2%, reflecting protocol differences.
  • Over 56% of Ethereum nodes are hosted in data centers, compared to 28% in Bitcoin, revealing greater hosting centralization in Ethereum.

Implications and Future Developments

The findings imply that permissionless blockchains diverge sharply in decentralization metrics, shaped by protocol design, network engineering, and economic forces. Bitcoin’s relay mechanisms engender lower orphan rates and more robust network dispersion, whereas Ethereum’s incentives facilitate the persistence of higher centralization in both hosting and mining. These results challenge simplistic narratives regarding blockchain decentralization, stressing the ongoing need for protocol innovation—including topology randomization, improved block propagation, and alternative consensus strategies—to address these limitations.

Looking ahead, the results motivate exploration of bandwidth admission control, peer selection enhancements, and dynamic mining reward structures. Further understanding and mitigation of the factors driving data center centralization are critical as networks scale. Finally, the study’s methodological framework establishes a foundation for future comparative analyses of emerging blockchain protocols and Layer 2 solutions.

Conclusion

This paper provides an in-depth quantitative assessment of decentralization properties in Bitcoin and Ethereum, revealing nuanced advantages and limitations in network, mining, and protocol design. The strong numerical evidence supports the conclusion that decentralization remains a multi-faceted challenge requiring continued research and engineering responses as blockchain systems evolve.

Paper to Video (Beta)

No one has generated a video about this paper yet.

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Collections

Sign up for free to add this paper to one or more collections.

Tweets

Sign up for free to view the 2 tweets with 2 likes about this paper.