Survey on Blockchain Networking: Context,

State-of-the-Art, Challenges

Survey on Blockchain Networking: Context,

State-of-the-Art, Challenges

MAYA DOTAN, Hebrew University, Israel

YVONNE-ANNE PIGNOLET, DFINITY, Switzerland

STEFAN SCHMID, Faculty of Computer Science, University of Vienna, Austria

SAAR TOCHNER and AVIV ZOHAR, Hebrew University, Israel

Blockchains, in general, and cryptocurrencies such as Bitcoin, in particular, are realized using distributed systems and hence critically rely on the performance and security of the interconnecting network. The requirements on these networks and their usage, however, can differ significantly from traditional communication networks, with implications on all layers of the protocol stack. This article is motivated by these differences and, in particular, by the observation that many fundamental design aspects of these networks are not wellunderstood today. To support the networking community to contribute to this emerging application domain, we present a structured overview of the field, from topology and neighbor discovery, over block and transaction propagation, to sharding and off-chain networks, also reviewing existing empirical results from different measurement studies. In particular, for each of these domains, we provide the context, highlighting differences and commonalities with traditional networks, review the state-of-the-art, and identify open research challenges. Our article can hence also be seen as a call-to-arms to improve the foundation on top of which blockchains are built.

CCS Concepts: • Computer systems organization → Embedded systems; Redundancy; Robotics; • Networks → Network reliability;

107

Additional Key Words and Phrases: Blockchains, payment networks, distributed computing

ACM Reference format:

Maya Dotan, Yvonne-Anne Pignolet, Stefan Schmid, Saar Tochner, and Aviv Zohar. 2021. Survey on Blockchain Networking: Context, State-of-the-Art, Challenges. ACM Comput. Surv. 54, 5, Article 107 (May 2021), 34 pages. https://doi.org/10.1145/3453161

1      INTRODUCTION

Blockchain technologies allow mistrusting entities to cooperate in the absence of a trusted third party,
 by implementing a secure distributed ledger that is collectively managed by a peer-to-peer network.

 Blockchain technologies have recently received much attention in the context of cryptocurrencies such as Bitcoin or Ethereum,
 by providing means to exchange digital assets relying on strong cryptography.

 In contrast to centralized digital currencies and central banking systems,
 cryptocurrencies offer decentralized control,
 and accordingly,
 much existing research in the blockchain field focuses on the underlying cryptographic primitives and on improved distributed blockchain protocols,
 e.

g.

,
 consensus.

The network required to connect the distributed system,
 however,
 has received relatively little attention.

 However,
 there is increasing evidence that the network can become the bottleneck and root-cause for some of the most pressing challenges cryptocurrencies specifically,
 and blockchains in general,
 face today.

 For example,
 the propagation of transactions and blocks (or other control messages for the execution of consensus algorithms),
 require unicast and multicast communication services.

 Blockchain miners that write transactions to the blockchain,
 are connected through dedicated miner peer-to-peer (P2P) networks [14],
 in addition to the public blockchain P2P network.

 Studies show that the cryptocurrency network layer is critical for scalability [24,
 63],
 security [45],
 and privacy [44] of a blockchain,
 and that an efficient network layer enables higher transaction throughput and stronger resilience against malicious actors [45].

 Networking issues are also not limited to overlays on the network layer.

 Besides node discovery and data routing,
 the provided network functionality for example includes the encoding and transmission of data,
 and error correction,
 as well as measurements of the network performance.

Interestingly,
 networks for blockchains and cryptocurrencies can differ significantly from traditional communication networks,
 also in terms of requirements and usage.

 For example,
 blockchain networks may come with different security requirements (e.

g.

,
 related to anonymity),
 may need to serve different traffic mixes (e.

g.

,
 more frequent broadcast of transactions and states of the blockchain),
 may need different routing mechanisms (e.

g.

,
 source routing),
 or may be more dynamic (e.

g.

,
 channels and fees in PCNs).

 Obviously,
 if in a cryptocurrency application no protocol participant can be trusted (not the users issuing transactions not the miners,
 and information providers may act maliciously),
 this requires a different design,
 and the incentives of all participants must be considered.

This article is a call-to-arms to the networking community to identify the unique requirements of blockchain networks and address the open issues.

Our Contributions.

 This article aims to provide a fast introduction to blockchain network issues with a focus on open research challenges.

 To this end,
 we provide a structured overview of blockchain networks,
 introducing the different aspects and their context,
 highlighting differences and commonalities with traditional networks,
 and reviewing the state-of-the-art.

 At the end of each section,
 we identify and discuss research questions.

 A main emphasis will be put on cryptocurrency applications specifically.

This article hence targets junior and senior researchers with different backgrounds (e.

g.

,
 in networking,
 algorithms,
 or game theory) who would like to get an overview of the state-of-the-art and start working in this area.

 It can also serve experts and decision-makers in the networking industry.

Related Work.

 This article surveys blockchain network issues with a focus on research challenges.

 The most closely related papers in this area are surveys and systematizations of knowledge efforts on cryptocurrencies and blockchains in general [16].

 Gudgeon et al.

 [49] provide a comprehensive survey of off-chain networks,
 and Neudecker et al.

 [81] survey the network layer

of permissionless blockchains,
 with a focus on attacks and the design space (but less on research questions,
 e.

g.

,
 revolving around incentives or mining).

 Katkuri presents a survey of data transfer and storage techniques in prevalent cryptocurrencies and suggests improvements [60].

 The focus is on aspects related to the broadcast networks underlying Ethereum,
 Nano and IOTA.

 Gervais et al.

 [45] give a thorough security analysis of proof of work blockchain systems; their focus is on the consensus layer (i.

e.

,
 block generation),
 whereas the network layer is abstracted.

 Troncoso et al.

 [105] show a broader perspective covering numerous systems apart from Bitcoin and Tor but also abstract from the network layer.

 A recent paper by Delgado-Seguara et al.

 [26] explores the characteristics of the P2P network established by Bitcoin,
 but abstracts from the design space of the network layer.

 Delgado et al.

 [26] provide an in-depth study of the Bitcoin P2P network.

 A preliminary and significantly shorter version of this article was presented in a conference [29]; the current version additionally covers offchain networks,
 networking and scalability aspects in sharding,
 as well as measurement studies,
 and includes more detailed and up-to-date discussions throughout the article.

Organization.

 The remainder of this article is organized as follows (see Figure 1).

 Section 2 provides an overview of some basic aspects of blockchain networks.

 These include incentives,
 topology,
 communication pattern and security.

 Next,
 we discuss core aspects,
 namely,
 block propagation,
 transaction propagation,
 P2P network topologies,
 sharding,
 off-chain networks,
 and empirical results in Sections 3–8.

 In each of these sections,
 we give some background information,
 present the state-of-the-art followed by open research questions that the writers of this article find interesting.

 Finally,
 in Section 9,
 we conclude this survey.