Celer Network Research and Analysis

Forward the Original Title：The Ultimate Off-chain Scaling Solution: Analyzing Celer Network

Video address

Today, we delve into an impressive project, the ultimate off-chain scaling solution where achieving TPS in the tens of millions is a breeze. Forget about Rollups and ZK, they’re in the shadows compared to the immense capability of Celer Network. Currently valued at $100 million, it holds the 199th position in the market.

Introduction

Celer Network is a high-performance Layer 2 scalability platform designed to bring internet-scale capabilities to blockchain through technological innovations. Its grand vision is to facilitate the commercialization and widespread adoption of blockchain by serving as a universal network system running on existing and future blockchains. Leveraging innovative off-chain scaling technology and incentive-aligned implicit cryptoeconomics, Celer Network delivers unprecedented performance and flexibility. In essence, it is dedicated to building a gateway platform for internet-scale blockchain applications using off-chain scaling technology, enabling everyone to develop, deploy, and use high-performance distributed blockchain applications quickly and conveniently.

Project Highlights

Off-chain scaling technology enables smart contract interactions among mutually distrustful entities off-chain, maintaining a multi-signature, immutable off-chain state machine. On-chain consensus is only invoked when absolutely necessary, such as when multiple parties cannot reach a consensus. Off-chain scaling is the only way to ensure the trustless and decentralized nature of blockchain while achieving horizontal scalability and privacy protection for distributed applications (dApps). It marks a turning point in the widespread adoption of blockchain technology and will serve as the engine and cornerstone for all scalable dApps.

Technical Overview

As a comprehensive platform that can be deployed on existing or future blockchains, Celer Network features a clear layered architecture that separates the complex off-chain platform into modular components. This architecture significantly reduces the complexity of system design, development, and maintenance, allowing each component to iterate easily and adapt to changes. A well-designed layered architecture should have open interfaces, encouraging different implementations on each layer as long as they support common cross-layer interfaces. Each layer only needs to focus on implementing its specific functionalities. Inspired by the successful layered design of the internet, Celer Network adopts an off-chain technology stack named cStack, which can be built on different blockchains. The cStack comprises the following layers in bottom-up order:

cChannel: Generalized state channel and sidechain suite

cRoute: Provably optimal value transfer routing

cOS: Development framework and runtime for off-chain enabled applications.

Celer’s architecture provides innovative solutions for all its layers. Below, we delve into the technical challenges and key features of cChannel, cRoute, and cOS.

cChannel

This layer is the closest to the underlying blockchain, directly interacting with the base public chain and providing real-time state updates to upper layers with a common abstraction structure within a limited timeframe. cChannel employs state channels and sidechain technology, which are the cornerstones of off-chain scaling platforms. State channels allow mutually distrustful parties to quickly reach consensus on the latest agreed-upon state in off-chain transactions and ensure their tamper-proof security through on-chain debt contracts. This concept was initially introduced by the Lightning Network to support high-throughput off-chain Bitcoin microtransactions. Since the inception of the Lightning Network concept, there has been research addressing various issues in the context of payment channel networks, such as routing and time-lock optimization. However, off-chain networks are still in the early stages of development, facing significant challenges in modularity, flexibility, and cost-effectiveness. cChannel addresses current challenges by introducing a series of new features:

• Generic off-chain state transition: Off-chain transfers can depend on DAG-based arbitrary state transitions, enabling Celer Network to support complex, high-performance offline dApps such as games, online auctions, insurance, prediction markets, and decentralized exchanges.
• Flexible and efficient value transfer: Multiple state channel and sidechain structures, diverse efficiency and tamper-proofing trade-off strategies, support rapid value transfer with general conditions, minimizing on-chain interactions and minimal fund lock-up.
• Purely off-chain contracts: Contracts not directly related to on-chain deposits require no on-chain operations or initialization unless a dispute is triggered. Each purely off-chain contract or object has a uniquely identifiable off-chain address, deployed on the blockchain only when necessary and allocated an on-chain address by the built-in off-chain address translator.

cRoute

Celer Network is a highly scalable dApps platform, and one of its most crucial promises is to support high-throughput value transfers on the platform. Off-chain value transfer is a fundamental requirement for many off-chain applications. While Celer Network has a broader vision than payment solutions, it has made groundbreaking improvements to off-chain payment routing, as it directly determines how much value can be transferred within the ecosystem and how quickly. All existing off-chain payment routes can be attributed to traditional “shortest path routing” algorithms, which may result in poor performance in off-chain payment networks due to fundamental differences in the link model. The link capacity in computer networks is stable and stateless (not affected by past transmissions). However, the link capacity in off-chain payment networks is stateful (determined by on-chain deposits and past payments), causing the network’s topology and link states to constantly change in a highly dynamic network. This makes it difficult for traditional shortest path routing algorithms to converge, resulting in low throughput, high latency, or even outages.

Celer Network’s payment routing module acknowledges this fundamental challenge, and cRoute introduces a Distributed Balanced Routing (DBR) that utilizes distributed congestion gradients. We highlight a few unique properties of DBR:

Provable Optimal Throughput: We prove that for any payment transaction request rate, if there exists a routing algorithm that can support that rate, DBR can always find that algorithm. According to our assessments, DBR increases throughput by 15 times and channel utilization by 20 times compared to state-of-the-art solutions.

Transparent Channel Balancing: Since the Lightning Network, maintaining channel balance has been an intuitive goal. However, existing attempts in channel balancing comprise heuristics that require heavy on-chain or off-chain coordination with poor guarantees. DBR integrates the channel balancing process with routing and maintains network balance without requiring any additional coordination.

Fully Decentralized: DBR is a fully decentralized algorithm where each node only needs to communicate with its neighboring nodes in the state channel network topology. The messaging cost in the DBR protocol is also low.

Failure Resilience: The DBR algorithm exhibits high fault tolerance; it can quickly detect and adapt to unresponsive nodes, supporting the maximum possible throughput on the remaining available nodes.

Privacy Preserving: Due to its multipath nature, the DBR algorithm naturally preserves privacy regarding the transmitted value without the need for additional privacy protection techniques (e.g., ZKSNARK). Importantly, the DBR algorithm can seamlessly integrate with onion routing to protect the anonymity of the source and destination.

cOS

Celer Network is a highly scalable dApps platform, and while on-chain dApps merely serve as a simple frontend connecting to the blockchain, off-chain dApps, despite their enormous potential for scalability, face challenges when building on traditional public chains. Celer Network introduces cOS, a development framework that allows everyone to easily develop, operate, and interact with scalable off-chain dApps, thereby reducing the additional complexity introduced by off-chain scaling. Celer Network enables developers to focus more on the logic of their applications and create optimal user experiences, while cOS handles cumbersome tasks, including:

Identifying dependencies between arbitrary off-chain and on-chain states.

Managing the tracking, storage, and dispute of off-chain states.

Tolerate intermediate node failures transparently.

Support multiple concurrent off-chain dApps.

Compile a unified implementation to different on-chain and off-chain modules.

Celer Network’s off-chain scaling solution is based on several key concepts:

State Channels: Celer Network uses state channel technology to achieve off-chain scaling. State channels are private communication channels established on the blockchain between two or more parties. Participants can conduct numerous transactions within these channels, with only the final settlement result being committed to the blockchain, significantly reducing the burden on on-chain transactions and improving efficiency.

Off-chain Computation: Celer Network allows off-chain participants to execute complex computational tasks without waiting for confirmation from the blockchain network. These computations may include smart contract execution, game logic, and data processing. Only the final results are submitted to the blockchain to ensure security and transparency.

Conditional Payment Network: Celer Network introduces a conditional payment network, enabling users to perform conditional payment operations within state channels. This means that payments can be associated with specific conditions, such as timestamps or event triggers, providing more possibilities for complex off-chain applications.

Network Interoperability: Celer Network’s scaling solution is designed to interoperate with different blockchain networks, such as Ethereum, Bitcoin, and others. This enables the transfer of assets and data between different blockchains, facilitating a more extensive ecosystem connection.

CELER’s implementation layer relies entirely on an economic design model, maintaining and operating the entire CELER network through Proof of Liquidity Commitment (PoLC), Liquidity Backing Auction (LiBA), and the State Guardian Network (SGN).

1. Proof of Liquidity Commitment (PoLC): Off-chain payments require open channels and a certain amount of token collateral. To address the challenge of financial capability not necessarily aligning with technical expertise, CELER introduces a Proof of Liquidity Commitment. Through asset locking mining, this mechanism ensures sufficient liquidity for off-chain service providers, enhancing off-chain asset liquidity. The longer the locking period, the more CELER tokens are obtained. Mining Reward Mechanism: Within a limited timeframe, Celer Network plans to incentivize Network Liquidity Backer (NLB) who locks a Collateral Commitment Contract (CCC) to support the system with an additional 20% CELR compared to the pre-mining allocation. The incentive is proportional to the Proof of Liquidity Commitment(PoLC) computational power for each NLB.
2. Liquidity Backing Auction (LiBA): In addition to addressing liquidity through locked mining, CELER employs auctions to tackle liquidity issues. Service providers can compete for liquidity through a “crowd lending” mechanism.
3. State Guardian Network (SGN): When users are not online, the State Guardian Network serves as a compact sidechain to protect off-chain states. CELER token holders can autonomously form the SGN and become State Guardians. Before going offline, a user can submit their state to the SGN by paying a fee, allowing the appointed guardians to protect the state for a certain period. Guardians are selected randomly based on the state hash and “responsibility score.”

The fundamental process of this auction mechanism involves the Borrower (OSP) submitting information on the required loan amount and loan duration to create a standard LiBA smart contract on the Celer Network and initiating the Liquidity Backing Auction (LiBA) process. This auction is conducted anonymously, with bidding information including the lender’s expected loan interest rate, the amount of collateralizable liquidity (for the duration specified by the auction initiator OSP), CELR token quantity, etc. To facilitate the participation of Proof of Liquidity Commitment (PoLC) miners (NLB) in the auction, the mechanism allows miners to contribute liquidity through the CCC contract. The auction process employs a VCG auction mechanism (complex details omitted here for brevity), calculating a “happiness index” based on bidders’ bid information during the auction. The final auction winner is selected according to certain rules, with a noteworthy point being that if bidders have similar scores, the one with a higher quantity of CELR tokens wins. After the auction concludes, the Borrower (OSP) is required to prepay interest to the LiBA auction contract. Upon receiving the prepaid interest, the auction contract generates IOU tokens in a 1:1 ratio and sends them to the state channel contract where the OSP resides (cCurrency, e.g., generating cETH against an ETH collateral). This IOU token serves as the medium for value transfer in off-chain state channels. Typically, before the expiration of the auction contract, OSP sends all borrowed cETH back to the auction contract for debt settlement. At this point, the auction contract becomes void. Since cETH and ETH have a 1:1 relationship, users can terminate off-chain transactions and withdraw funds without risk at any time. (3) SGN (Enhancing Celer Network’s Availability via SGN): SGN serves as an off-chain state guardian network, acting as a sidechain to safeguard the off-chain states of offline clients. CELR token holders can become off-chain state guardians by pledging tokens. Guardians for submitted off-chain state tasks are randomly selected based on the state hash and “responsibility score.” The more tokens pledged, the higher the likelihood of being delegated to guard off-chain state tasks, leading to increased fee earnings. Users can pay a fee to submit off-chain states to the SGN network for a period before going offline, preventing unfavorable transaction revocation, among other benefits. In summary, LiBA and PoLC focus on introducing the intermediate steps of the state interaction process off-chain more easily, while SGN concentrates on protecting the off-chain state transition process, aiming to bring back off-chain states to the blockchain when necessary (e.g., in case of potential fraud by the counterparty). In the entire economic system, CELR tokens can be used for collateral in liquidity lending anti-fraud bonds, payment mediation for channel registration fees, transaction fees, and other possible service fees. Additionally, in the first 5 years of system operation, new CELR tokens will be generated through PoLC mining, and LiBA only requires tokens to be pledged, with the CELR belonging to the borrower after the pledge period. After the 5-year PoLC mining period, LiBA will start consuming CELR, and the consumed CELR will no longer be returned to the borrower but injected into the system as continuous PoLC mining rewards. In terms of security, Celer has the following designs:

1. NLP tends to choose OSPs with no default records for lending.
2. Newly minted cETH is only allowed to transfer to state channel contracts in a whitelist.
3. cETH can only be gradually spent at a capped rate.
4. OSPs can undertake various measures to maintain a secure infrastructure, such as compartmentalized multi-node deployment, formal verification of security access rules of network infrastructure and more.

Team

The architecture and off-chain incentive design of the entire Celer Network are explained above, and overall, the approach is quite clear. However, it requires strong implementation capabilities, making the team composition crucial. Fortunately, the Celer team is formidable and consists of known individuals. Here’s an overview of key team members:

Founder, Mo Dong: Graduated from Shanghai Jiao Tong University and received a PhD in CS from UIUC in 2017. He was a founding member of Veriflow, a network formal verification security company, where he served as engineering team leader and product manager. Dr. Dong is an early participant, developer and pioneer in distributed technology and blockchain circles. The distributed system and network protocol formal verification software developed earlier has been deployed in the world’s top 50 companies. In 2017, he started teaching blockchain smart contract development courses, cultivating a large number of developers for the community.

1. Core Members

Junda Liu: Graduated from Tsinghua University, and earned a Ph.D. in CS from UC Berkeley in 2011. He has 7 years of experience at Google, and pioneered the use of DAG for routing, achieving 1,000 times better network repair capabilities compared to industry standards.

Xiaozhou Li: Graduated from Tsinghua University, and earned a Ph.D. in CS from Princeton. He has 2 years of experience at Barefoot Networks.

Qingkai Liang: Graduated from Shanghai Jiao Tong University, and earned a Ph.D. in CS from MIT.

1. Operations Team (weight 45%, score 42.75)

Pengying Wang: Graduated from Boston University, and serves as Vice President of MIT-CHIEF (Massachusetts Institute of Technology-China Innovation and Entrepreneurship Forum).

Sirong Li: Holds a Master’s degree in Business Analytics from the University of Rochester, with work experience in Fortune 500 companies with successful marketing cases. The team consists of 12 members, including 9 technical developers.

1. Investment Advisors:

Christos Kozyrakis:

Professor of Electrical Engineering and Computer Science at Stanford University.

Alan Mishchenko: Full-time researcher and engineer at UC Berkeley. Holds a Ph.D. in Computer Science from Glushkov Institute of Cybernetics in Ukraine.

Shousheng Zhang: Physics professor at Stanford University. Founder and Chairman of Danhua Capital.

Tokenomics

The project was launched in 2019 with a maximum supply of 10,000,000,000 CELR. Currently, 7,743,424,107 CELR is in circulation, representing a circulation rate of 77.4%. The current token price is $0.013, with the highest point being $0.1984 on September 26, 2021. This indicates a decline of over 90%. In terms of token distribution, the team holds 18.3%, and the foundation holds 17%, totaling 35%, which is a relatively high proportion. The Total Value Locked (TVL) on cBridge is $88 million. Despite the mainnet not being launched yet, achieving this volume is considered respectable.

In conclusion, this project presents innovative ideas as a solution for off-chain scalability. Positioned as an off-chain scalability solution, it introduces new technical solutions and an economic model, demonstrating significant performance improvements through experimental simulations. The development team showcases notable strength, guided by renowned researchers from prestigious institutions and experienced developers serving as advisors. There is potential for groundbreaking advancements in off-chain scalability with this project. However, it’s crucial to acknowledge the inherent limitations of state channels, which are more suitable for high-frequency interactions among multiple users and less applicable to low-frequency interactions with frequently changing users. Despite these challenges, the success of off-chain state channels as an enhancement suite for public blockchains could contribute to the practicality of blockchain technology. While the project faces significant implementation challenges, it’s important to note that current Transaction Per Second (TPS) requirements may not necessarily demand extremely high values. Existing high-speed public chains, including Layer 2 solutions, can adequately meet current needs. Nevertheless, this project presents a well-thought-out solution, and its current market valuation is relatively low compared to other L2 solutions. As the project has not yet been fully realized, there exists a potential undervaluation. Once the mainnet is launched successfully, it is likely to experience a substantial surge. Therefore, it might be worthwhile to include this project in our watchlist.

Disclaimer：

1. This article is reprinted from[书中自有大饼屋], Forward the Original Title‘ The Ultimate Off-chain Scaling Solution: Analyzing Celer Network’,All copyrights belong to the original author [朱老师321]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.