off-chain Scalability Depth Analysis

Authors: Ellaine Xu, Hettie Jiang, June Wang, Walon Lin, Yiliu Lin

1. The Necessity of Scalability

The future vision of blockchain is decentralization, security, and scalability, but usually only two of these can be achieved, which is known as the impossible triangle problem of blockchain. For many years, people have been exploring how to increase the throughput and transaction speed of blockchain while ensuring decentralization and security, that is, to solve the scalability issue.

Let us first define the decentralization, security, and scalability of blockchain:

• Decentralization: Anyone can become a node and participate in the blockchain system. The more nodes there are, the higher the degree of decentralization.
• Security: The higher the cost of gaining control of the blockchain system, the higher the security, making it resistant to a larger proportion of attacks.
• Scalability: The ability of blockchain to handle a large number of transactions.

The first major hard fork of the Bitcoin network originated from scalability issues. Since 2015, there has been a division in the Bitcoin community regarding scalability, with one side supporting larger blocks and the other side advocating for the Segregated Witness solution. On August 1, 2017, the side supporting larger blocks began running a new client system, resulting in the first major hard fork in Bitcoin's history, which gave rise to the new cryptocurrency BCH.

The Ethereum network also chose to sacrifice some scalability to ensure the security and decentralization of the network. From the CryptoKitties of 2017 to the rise of DeFi, GameFi, and NFTs, the market's demand for throughput has been constantly increasing, but Ethereum can only process 15-45 transactions per second, leading to increased transaction costs and longer settlement times. The ideal scaling solution is to maximize the transaction speed and throughput of the blockchain network without sacrificing decentralization and security.

2. Types of Scalability Solutions

We classify the scaling solutions into two main categories: on-chain scaling and off-chain scaling based on the criterion of "whether to change a layer of the mainnet."

2.1 on-chain scalability

Core concept: a solution to achieve scalability by changing a layer of the mainnet protocol, with the current main solution being sharding.

There are various solutions for on-chain scaling, and the following briefly lists two of them:

• Option one is to expand the block space, which means increasing the number of transactions packed in each block, but this will raise the requirements for nodes and reduce the level of decentralization.
• Option two is sharding, which divides the blockchain ledger into several parts, with different shards responsible for different accounting. This can reduce node pressure, but it will lower the overall security of the network.

Changing a layer of the mainnet protocol may have unpredictable negative effects, as any security vulnerabilities at the underlying level could severely threaten the overall security of the network.

2.2 off-chain scaling

Core concept: a scaling solution that does not change the existing Layer 1 mainnet protocol.

The off-chain scalability solutions can be further divided into Layer 2 and other solutions:

• Layer2: State Channels, Plasma, Rollups
• Other solutions: Sidechains, Validium

3. Off-chain Scaling Solutions

3.1 State Channels

3.1.1 Overview

State channels stipulate that users need to interact with the mainnet only when the channel is opened, closed, or when disputes are resolved, allowing interactions between users to take place off-chain to reduce transaction costs and achieve unlimited transaction frequency.

State channels are simple P2P protocols suitable for turn-based applications, such as two-player chess games. Each channel is managed by a multi-signature smart contract running on the mainnet, which controls the assets deposited in the channel, verifies state updates, and arbitrates disputes between participants.

3.1.2 Timeline

• 2015/02: Joseph Poon and Thaddeus Dryja released the draft of the Lightning Network white paper.
• 2015/11: Jeff Coleman first systematically summarized the concept of State Channel
• 2016/01: The Lightning Network white paper was officially published
• 2017/11: Proposed the first State Channel design specification Sprites based on the Payment Channel framework.
• 2018/06: Counterfactual proposed a detailed design for Generalized State Channels.
• 2018/10: Proposed the concepts of State Channel Networks and Virtual Channels
• 2019/02: The concept of state channels expanded to N-Party Channels
• 2019/10: Pisa expanded the Watchtowers concept to solve the problem of participants needing to be online continuously.
• 2020/03: Hydra proposed Fast Isomorphic Channels

3.1.3 Technical Principles

Workflow of State Channels:

1. Alice and Bob open a state channel by depositing funds into the mainnet contract and signing to confirm.
2. Alice and Bob can conduct unlimited off-chain transactions, with each transaction requiring signatures from both parties.
3. If Alice wants to close the channel, she needs to submit the final state to the contract. If Bob signs to approve, the contract will immediately execute the distribution of funds; if Bob does not respond, she will have to wait until the "challenge period" ends to receive the funds.

Workflow in a pessimistic scenario:

If Bob does not respond to Alice's state update signature, Alice can challenge by submitting her last valid state to the contract. Bob can submit the next state in response within a certain period; if he does not respond, the contract will automatically close the channel and return the funds to Alice.

3.1.4 Advantages and Disadvantages

Advantages:

• Instant Confirmation
• High Throughput
• Good privacy
• Strong versatility

Disadvantages:

• Requires pre-locking of funds
• Participants must remain online
• Support for fewer participants
• Unable to solve the large-scale exit problem.

3.1.5 Application

Bitcoin Lightning Network:

• Proposed in 2015, mainnet version released in 2018
• In November 2022, there were 76,236 payment channels with channel funds of 5049 BTC.
• The ecosystem covers multiple categories including payment, wallet, node management, and more than 100 applications.

Ethereum Lightning Network:

• Established in 2017, mainnet version released in 2020
• High barrier to entry, slow ecological development
• Transition to operating on Layer2 Rollup networks

Celer Network:

• Increased the incentive layer of the Lightning Network
• Suitable for high-frequency interactive applications such as esports platforms
• Launched cross-chain bridge products like cBridge.

3.2 Sidechains

3.2.1 Overview

Sidechains are a form of blockchain that emerged to accelerate Bitcoin transactions and can utilize more complex contracts or improve consensus mechanisms. The transaction results of sidechains will ultimately be recorded on the validator's end and transmitted back to the main chain.

3.2.2 Timeline

• 2012/01: The concept of Bitcoin sidechains was first proposed
• 2014/10: Publication of the Bitcoin sidechain paper
• 2017/04: POA Network testnet launched
• 2017/10: Matic Network launched
• 2018/01: Skale testnet goes live
• 2020/06: Matic PoS Chain mainnet launch
• 2021/02: Ronin mainnet launched
• 2021/12: xDai Chain merged with Gnosis Dao to form Gnosis Chain

3.2.3 Technical Principles

There are mainly two ways for sidechains to communicate with the main chain:

1. Bidirectional Anchoring ( Symmetric Pegged ): Validators on the main chain and side chain record each other's current state in real-time, using SPV technology to verify transactions.

2. Incoherent anchoring ( Asymmetric Pegged ): Sidechain validators monitor mainchain activities, but the mainnet cannot confirm the state of the sidechain, requiring the introduction of a Certifiers mechanism to validate sidechain transactions.

More and more sidechains are currently choosing to use third-party notaries ( PoA ) mechanisms or Relayers middleware to confirm the main chain block status.

Summary of sidechain mechanism:

• Assets from the main chain to the side chain: the main chain locks assets, and the side chain generates wrapped assets.
• Assets from side chain to main chain: side chain destroys wrapped asset, main chain unlocks asset

3.2.4 Advantages and Disadvantages

Advantages:

• Highly customizable
• Supports complex smart contracts
• Cross-chain asset transfer
• Independent governance mechanism

Disadvantages:

• lower security
• Centralization Risk
• The complexity of cross-chain communication is high
• Bi-directional anchoring requires a longer confirmation time.

3.2.5 Application

xDai Chain( now Gnosis Chain):

• Mainnet launched in September 2018, merged with Gnosis in April 2022
• Adopting PoSDAO consensus mechanism
• TVL is approximately $53 million, with 35 ongoing projects in the ecosystem.

Polygon:

• The mainnet of Matic PoS Chain and Plasma Chain went live in June 2020.
• In 2021, the brand was upgraded to Polygon and transitioned to an aggregator.
• There are over 37k Dapps, 1.8B total transactions, and over 135M users.

Ronin:

• The sidechain developed for the Axie Infinity game
• The mainnet went live in March 2021, using the PoA consensus mechanism.
• In March 2022, suffered a hack attack of 624 million USD.

3.3 Plasma

3.3.1 Overview

Plasma is a framework for building scalable Dapps, designed to minimize user trust in off-chain Operators. The basic principle of Plasma is that even if the Plasma chain experiences a security failure, all users' assets can still be withdrawn from the Plasma chain and returned to the mainnet.

3.3.2 Timeline

• 2017/08: Plasma white paper released
• 2018/01: Proposed the first official Plasma application Plasma MVP
• 2018/03: Proposed Plasma Cash to solve the large-scale exit problem
• 2018/06: Proposed Plasma Debit
• 2018/11: Proposed Plasma Prime
• Since 2019: The Ethereum community began exploring Rollups solutions.

3.3.3 Technical Principles

The core idea of Plasma:

• off-chain execution: Most of the work is handled outside the mainnet.
• State commitment: Store a compressed version of the Plasma chain state using Merkle Root
• Exit mechanism: Users must prove to the mainnet that they have withdrawable funds and that the amount is correct.

Usage process:

1. Users deposit funds into the mainnet Plasma contract.
2. Users trade on the Plasma chain
3. The operator packages the transaction and submits the state commitment to the mainnet.
4. Users can withdraw funds after initiating a withdrawal request and going through the challenge period.

3.3.4 Advantages and Disadvantages

Advantages:

• High throughput
• Low transaction fees
• Inherit the security of the mainnet

Disadvantages:

• Complex exit mechanism
• Data availability issues
• Users need to monitor frequently
• Function limited

3.3.5 Application

Main applications:

• Plasma Group → Optimism(Optimistic Rollup)
• OMG Network → Boba Network(Optimistic Rollup)
• Polygon( previously Matic Network) → Full-stack L2 solution

Summary: Plasma is a technical transition solution, constrained by issues inherent to the technology itself, and most applications have quickly shifted towards the development of Rollups-based solutions.

3.4 Rollups

3.4.1 Overview

The core idea of Rollups is to place the computation process and state storage off-chain, while storing state commitments and compressed transaction data on the main chain.

Rollups are divided into two types:

• Optimistic Rollups: Ensure correctness through fraud proofs.
• ZK Rollups: Ensure correctness through effective proof ( zero-knowledge proof )

3.4.2 Technical Principles

Core Mechanism:

• off-chain virtual machine executes transactions and state storage
• Upload the compressed transaction data and state root to the main chain contract