Source-linked AI summary
A Survey of Layer-Two Blockchain Protocols
Ankit Gangwal, Haripriya Ravali Gangavalli, Apoorva Thirupathi
TL;DR
Blockchain scalability is limited by low throughput and high latency, while Layer-1 solutions require changes to fundamental blockchain attributes. This paper surveys Layer-2 protocols by building a broad taxonomy, analyzing protocol classes and concerns, and comparing their properties. It concludes that Layer-2 protocols improve scalability with differing security assumptions and guarantees.
Problem
Blockchain systems suffer from low transaction rates and high processing latency, while rapid Layer-2 development has produced a fragmented literature lacking a comprehensive, systematic guide.
Method
The paper surveys Layer-2 protocols, creates a broad taxonomy, explains protocol classes and implementations, and examines networking, security, privacy, and comparative feasibility.
Results
The study consolidates and systematizes Layer-2 protocols, detailing their characteristics, limitations, issues, and differing security assumptions and guarantees.
Takeaways & Limitations
The survey provides a comprehensive guide for understanding and comparing Layer-2 scalability solutions and their practical concerns.
Abstract
from arXiv · showhide
After the success of the Bitcoin blockchain, came several cryptocurrencies and blockchain solutions in the last decade. Nonetheless, Blockchain-based systems still suffer from low transaction rates and high transaction processing latencies, which hinder blockchains' scalability. An entire class of solutions, called Layer-1 scalability solutions, have attempted to incrementally improve such limitations by adding/modifying fundamental blockchain attributes. Recently, a completely different class of works, called Layer-2 protocols, have emerged to tackle the blockchain scalability issues using unconventional approaches. Layer-2 protocols improve transaction processing rates, periods, and fees by minimizing the use of underlying slow and costly blockchains. In fact, the main chain acts just as an instrument for trust establishment and dispute resolution among Layer-2 participants, where only a few transactions are dispatched to the main chain. Thus, Layer-2 blockchain protocols have the potential to transform the domain. However, rapid and discrete developments have resulted in diverse branches of Layer-2 protocols. In this work, we systematically create a broad taxonomy of such protocols and implementations. We discuss each Layer-2 protocol class in detail and also elucidate their respective approaches, salient features, requirements, etc. Moreover, we outline the issues related to these protocols along with a comparative discussion. Our thorough study will help further systematize the knowledge dispersed in the domain and help the readers to better understand the field of Layer-2 protocols.
I. INTRODUCTION
Blockchain scalability remains constrained by low throughput and high latency, while Layer-1 approaches require changes to core blockchain design. Layer-2 protocols address this by moving transactions off-chain while retaining the main chain for trust and dispute resolution.
- Scalability challenge: Roughly 10 TPS and one block every ten minutes illustrate Bitcoin’s scalability limitations relative to conventional payment systems.Limited block capacity and consensus-driven block intervals restrict transaction throughput and increase latency.
- Layer-1 approaches: Layer-1 solutions scale blockchains by modifying block data, changing consensus, sharding networks, or using Directed Acyclic Graphs.Because these approaches alter core blockchain elements, they commonly lack backward compatibility and can create implementation issues such as forks.
- Layer-2 approach: Layer-2 protocols scale blockchains without altering the underlying consensus mechanism or Layer-1 trust assumptions.They are built above the blockchain stack and use the main chain selectively rather than changing its core design.
- Layer-2 approach: Off-chain execution reduces main-chain transaction load while preserving backward compatibility, with security provided by collateral or delayed finality.Payment channels use collateral, whereas commit chains use delayed finality.
- Survey scope: The survey organizes a fragmented Layer-2 literature through a broad taxonomy, detailed protocol analysis, and comparative discussion of feasibility, security, privacy, and networking.It aims to help readers assess diverse Layer-2 solutions and their respective advantages, limitations, requirements, and concerns.
A. Blockchain and HTLC
Blockchain systems rely on consensus, access controls, and scripting to maintain shared state, while HTLCs provide hash- and time-based conditions for trust-minimized payments. The paper situates these mechanisms within a broader effort to systematize blockchain scalability research and Layer-2 protocols.
- Blockchain fundamentals: Blockchains are append-only chains whose participants use consensus to agree on a shared state and maintain ledger integrity.They may be permissionless or permissioned, and their scripting language determines expressiveness.
- Blockchain fundamentals: Underlying blockchain expressiveness affects Layer-2 protocol design, with Bitcoin using a Turing-incomplete script and Ethereum supporting Turing-complete smart contracts.Layer-2 protocols can be built on both permissioned and permissionless blockchains.
- HTLC: HTLCs use hash verification and time expiration as a trust-based construction for payment channels and several Layer-2 protocols.The Lightning Network is given as a representative implementation.
- HTLC: In a payment channel route from A through B to C, the receiver’s secret hash-lock propagates backward, while decreasing timeouts let intermediaries claim funds conditionally.C generates secret Z, A and B lock funds using Hash(Z), and C can claim funds by producing Z.
- Related works: The survey addresses a fragmented Layer-2 literature through a broader taxonomy and coverage of protocol characteristics, concerns, and scope through December 2021.It was designed as a comprehensive guide because prior surveys omitted major protocol classes.
III. LAYER-TWO BLOCKCHAIN PROTOCOLS
Layer-Two protocols organize off-chain scalability approaches into channels, side/child chains, cross chains, and hybrid solutions. State channels exchange application states off-chain and commit the final state to the main chain, using several replacement and dispute techniques.
- Taxonomy: Layer-Two solutions are broadly categorized into channels, side/child chains, cross chains, and hybrid solutions.The taxonomy covers both Layer-1 and Layer-2 blockchain scalability solutions.
- State channels: State channels exchange arbitrary application states among participants and commit the final channel state to the main chain through a smart contract.They typically use multisignature contracts and timelock instructions, with states exchanged among channel participants.
- State channels: Frequent off-chain state exchange makes state channels faster than on-chain exchanges and improves the parent blockchain’s transaction rate.Their lifecycle comprises establishment, execution, and termination phases, beginning with funds or assets locked through a main-chain contract.
- State replacement: State replacement techniques include Replace-by-Incentive, Replace-by-Timelock, Replace-by-Revocation, and Replace-by-Version.RbI and RbT insert the latest state once, whereas RbR and RbV support invalidation through counter-evidence and dispute processes.
- Operational requirements: Channel-based Layer-Two solutions initially require participants to stay online, although watchtowers can monitor transactions and disputes while participants are offline.Watchtower services impose additional costs on participants.
2) Payment channels:
Payment channels specialize state channels for payments, enabling repeated off-chain transfers while reducing on-chain activity and fees. Their direct-channel and fund-locking requirements motivate channel factories, payment-channel networks, hubs, and virtual channels.
- Payment channels: Payment channels tailor state channels to payment applications and evolved from one-way to bidirectional channels.Their lifecycle includes establishment, execution, and termination or dispute.
- Limitations: Payment-channel setup requires funds locked exclusively to the channel, main-chain confirmation, and a dedicated channel between participants.These requirements constrain payment-channel use for micropayments.
- Channel extensions: Channel factories let n participants jointly lock an n-party deposit and reallocate it to create pairwise payment channels without separately funding each pair.Participants update the shared deposit when establishing a direct channel.
- Channel extensions: Payment-channel networks route transactions through intermediary channels, such as A→B→C, using incentives and conditional payment constructions like HTLCs.The conditional transactions must be atomic so the routed payment resolves consistently.
- Channel extensions: Payment-channel hubs use star topologies to relay payments, reducing routing overhead and funds locked by individual nodes in payment-channel networks.Interconnected hubs can reduce routing length, routing cost, and collateral cost at each channel.
- Channel extensions: Virtual channels create the appearance of a direct payer-payee channel while intermediaries lock funds for a fixed duration.They reduce the need for intermediaries to actively participate in related transactions.
B. Side/Child chains
Side chains are independent ledgers operating alongside a main chain, using their own consensus and a two-way peg to offload computation and transfer assets. They offer faster processing and persistent participation, while custodial and non-custodial designs differ in where assets and states are secured.
- Side chains: A side chain is an independent distributed ledger running in parallel with the main chain.It offloads computationally heavy work and can transfer assets across blockchains.
- Side-chain operation: Side chains generally use their own consensus mechanisms and communicate with the main chain through a two-way peg.Synchronization includes a main-chain confirmation period and a contest period for preserving fund-conversion integrity.
- Fund movement: Funds can move within a side chain without interacting with the main chain, while remaining bonded to the parent chain.Redeeming funds requires locking them to a special output and using a corresponding SPV proof on the main chain.
- Advantages: Independent consensus and controllable block parameters make side-chain transactions typically faster than main-chain transactions and reduce main-chain load through offloading.Side chains can run permanently, admit new participants, and confine compromise or damage to the side chain.
- Classification: Side chains are classified as custodial or non-custodial according to their asset-security model.Non-custodial side chains secure assets and states through main-chain smart contracts and include commit or Plasma chains and rollups.
1) Commit chains:
Commit chains move transactions off-chain under a non-custodial operator, while smart contracts and the parent chain support security and dispute handling. Rollups similarly batch compressed transactions for on-chain verification, with Optimistic and zk designs trading verification assumptions and costs.
- Commit chains: Commit chains use a non-custodial operator to maintain off-chain accounts, while a smart contract prevents operator misbehavior.
- Commit chains: Recipients can receive funds while offline, and commit chains provide eventual finality without operator collateral or instant finality when collateral is staked.
- Commit chains: Commit chains inherit the parent chain’s consensus security but retain operator and data-maintenance limitations for challenging states and exiting.
- Rollups: Rollups compress off-chain transaction batches and retain a Merkle root on-chain, enabling verification and faster withdrawals while reducing main-chain load.
- Rollups: Optimistic Rollups assume validity unless challenged with fraud proofs, whereas zk Rollups attach validity proofs verified on-chain.
- Rollups: Optimistic Rollups suit general-purpose computation but have limited throughput, while zk Rollups fit simple payments but incur high proof-computation cost and complexity.
C. Cross Chains
Cross chains address interoperability by enabling asset transfers and communication among independent blockchains, while Layer-2 hybrid protocols reduce dispute-resolution dependence or peer-trust requirements. The survey describes architectures including Celer, notary schemes, and blockchains of blockchains.
- C. Cross Chains: Cross chains transfer assets between independent blockchains and establish mutual trust procedures despite differing consensus mechanisms.
- C. Cross Chains: Celer Network combines generalized state channels, routing, and coordination between on-chain and off-chain states to support rapid off-chain transitions.
- C. Cross Chains: Notary schemes use an entity that observes multiple blockchains and creates a transaction on one chain when a corresponding event occurs on another.
- C. Cross Chains: A blockchain of blockchains uses a backbone chain to let independent blockchains share data or tokens while supporting customized application-specific networks.
- C. Cross Chains: Cosmos organizes independent blockchains as zones connected through hubs, reducing the number of connections required for inter-zone communication.
- Hybrid solutions: Hybrid Layer-2 solutions either move part of dispute resolution off-chain through bisection protocols or use secure execution to remove peer-trust requirements.
2) TEE-based solutions:
TEE-based solutions use trusted execution environments to protect off-chain execution and remove on-chain collateral used to establish trust. Examples support payment channels, asynchronous transactions, replicated treasury states, and client privacy, but depend on TEE integrity.
- 2) TEE-based solutions: TEEs protect the integrity and confidentiality of loaded data, allowing blockchain-scaling protocols to replace on-chain collateral with a mutually trusted execution environment.
- 2) TEE-based solutions: Teechan enables mutually distrusting nodes to exchange funds peer-to-peer while TEEs securely maintain and update channel state without the parent blockchain.
- 2) TEE-based solutions: Teechain executes transactions asynchronously with the main chain and uses TEE-protected treasuries plus replicated committee states to handle treasury failures.
- 2) TEE-based solutions: Tesseract, BITE, and ZLiTE apply TEE-based designs to cryptocurrency exchange or lightweight-client privacy use cases.
- 2) TEE-based solutions: All TEE-based solutions rely on TEE integrity, while TEEs have their own vulnerabilities and concerns.
2) Re-balancing:
The section covers payment-channel rebalancing, watchtower-assisted availability, privacy and routing concerns, and attacks that exploit intermediary knowledge or channel conditions. These mechanisms aim to preserve usability but introduce operational and security trade-offs.
- 2) Re-balancing: REVIVE safely rebalances skewed payment channels using funds in other channels and an elected untrusted leader that coordinates participants’ requests.
- Availability: Watchtowers let offline participants delegate transaction monitoring and dispute detection, receiving fees while facing penalties for failing to report disputes.
- Stability and privacy: A watchtower may be bribed when the bribe exceeds its collateral, creating a route to dishonest behavior against customers.
- Stability and privacy: Payment routing can expose payer, payee, or channel-capacity information, allowing protocols that learn capacities over time to bypass stated privacy notions.
- A. Wormhole attack: Wormhole attacks use two malicious path nodes to bypass intermediate channels and steal rewards from benign intermediaries.
B. Flood and loot
Flood and loot is a systemic Lightning attack that uses attacker-controlled channels and unresolved HTLC payments to force victims toward on-chain fund recovery. This simultaneously overloads the blockchain and creates an opportunity for theft.
- Attack setup: Flood and loot targets Lightning by opening multiple channels from an attacker-controlled source through victim nodes to an attacker-controlled destination.A victim is exposed only if it opens a channel with the attacker’s source node.
- Victim response: As HTLC timeouts approach, victims must close channels and claim their locked funds on the parent blockchain.Victims participating in multiple source-to-target paths face a worse exposure.
- Impact: Simultaneous channel closures benignly overload the blockchain, creating a window for the attacker to steal funds.The attack converts off-chain payment activity into concentrated on-chain recovery demand.
- Attack execution: The attacker sends multiple HTLC payments through these channels, accepts them at the destination, and refuses to resolve them at the source.The payments are designed to maximize channel usage and funds locked in unresolved HTLCs.
C. Griefing attack
Griefing attacks stall payment networks by exhausting channel capacity rather than directly stealing funds. A representative multihop attack locks progressively smaller amounts across timed off-chain contracts until participants must resolve or refund on-chain.
- Purpose and impact: Griefing attacks stall payment networks by exhausting channel capacity, causing temporal fund loss, lower throughput, and routing disruption.Their objective is stalling rather than directly stealing funds or information.
- Representative setup: In a path A→D→C→B, each hop locks funds and charges a processing fee under progressively shorter contract deadlines.A locks p + 2p′ for t + 2t′, D locks p + p′ for t + t′, and C locks p for t.
- Resolution: The final recipient must resolve the payment before its deadline; otherwise, the preceding node can refund on-chain by closing its channel.The same resolution pattern propagates backward through the route.
- Related attack: Time dilation attacks eclipse victims, delay block delivery, and leave them operating with outdated network information.The attacker occupies every victim connection with pseudonymous nodes and delays forwarding received blocks.
- Limitations and defenses: A dilated victim can lose races to finalize contracts before expiration, while detection and mitigation remain difficult.Watchtowers provide substitute defense but depend on assumptions about honesty and efficiency.
E. Balance lockdown attack
Balance lockdown attacks block victims’ funds so they cannot route further payments, giving attackers a competitive position. Their effectiveness is characterized by attack effort relative to blocked capacity and by the duration of the lock.
- Attack objective: Balance lockdown attacks block a victim node’s funds, preventing further routing and reducing system efficiency.The attack can block selected payment paths and confer a dominant position on the attacker.
- Mechanism: Because multihop payments are atomic, an attacker can lock amount p on every intermediate victim by sending a payment through the path.Intermediate nodes must retain funds until the complete path is established.
- Effectiveness metrics: Attack Effort Ratio (AER) equals the capacity required to launch the attack divided by the capacity it effectively blocks.Transaction cost and completion time determine attack feasibility.
- Effectiveness metrics: Total Blocked Time (TBT) measures how long victims’ funds remain blocked during the transaction.Countermeasures seek to increase AER and reduce TBT, although regulating TBT trades off against AER.
- Related privacy attack: Balance discovery attacks disclose individual channel balances despite publicly known total capacity, compromising payment privacy.The last successful payment approximates one side’s balance, while the other follows by subtracting from total capacity.
- Related availability attack: Congestion attacks block multiple routes by making the attacker both payment source and destination, allowing repeated delays until HTLC expiry.The attacker can relaunch the attack after cancellation, potentially blocking paths for weeks.
VI. DISCUSSION
The comparison shows that Layer-2 designs exchange different performance, capital, finality, liveness, contract, and privacy properties. The survey presents these trade-offs while noting that scalability comes with differing security assumptions and guarantees.
- Performance and capital efficiency: Channels provide almost instant Layer-2 transfers at low transaction cost but require collateral locked on-chain.Other Layer-2 solutions are therefore perceived as more capital efficient than channels.
- Withdrawal and finality: Channel withdrawals require one on-chain confirmation, whereas commit chains and Optimistic Rollups typically delay withdrawals for security and dispute resolution.zk Rollup withdrawals can take fewer minutes in the best case.
- Withdrawal and finality: Channels offer instant finality with full security guarantees, while other Layer-2 solutions typically use delayed finality.Other solutions can present instant user confirmation without providing the same full security guarantees.
- Cryptography and liveness: Most Layer-2 protocols use standard cryptography, while SNARKs and STARKs are heavily used by zero-knowledge protocols.Commit chains support offline transaction receipt, although users are advised to inspect checkpoint commitments periodically.
- Smart contracts and privacy: Optimistic and zk Rollups support importing existing EVM bytecode with minor modifications, enabling flexible smart-contract support.Only zk Rollups address transaction deanonymization and user profiling by default.
- Survey scope: The survey’s taxonomy and comparison organize Layer-2 protocols while relating scalability benefits to differing security assumptions and guarantees.It discusses protocol classes, implementations, associated issues, and comparative properties.