# Survey Liquidity Solutions for Fiber (Fleeting) ## Metadata **Status**:: #x **Zettel**:: #zettel/fleeting **Created**:: [[2025-09-18]] ## Synopsis - Liquidity swaps, splicing, and rebalancing - Loop payment - Lightning pool - Channel factories: open multiple channels using a single on-chain tx - The use of Wumbo channels (high-capacity channels beyond the original limits) and collaborative channel management strategies are being actively studied and deployed for enterprise and high-volume use cases. - Overpayments https://delvingbitcoin.org/t/multichannel-and-multiptlc-towards-a-global-high-availability-cp-database-for-bitcoin-payments/1983/ - https://lightning.engineering/posts/2020-11-02-lightning-pool/ - https://www.btcstudy.org/2024/03/01/lightning-network-technology-improvement-and-users-experience-part-5/ ## Submarine Swap - Bob generates a secret value R (preimage) and its hash H. - Bob creates an HTLC on the Bitcoin blockchain using hash H: he will pay Carol 10,000 satoshis, provided he can reveal secret R within 5 blocks; otherwise, the funds will return to Bob. - Carol creates an HTLC in his payment channel with Bob using the same hash H: he will pay Bob 10,000 satoshis from the channel, provided he can reveal secret R within 4 blocks; otherwise, the funds will return to Carol (for simplicity, we are not considering any service fees charged by the exchange service). - Bob uses the secret R to unlock the HTLC in the channel and takes 10,000 satoshis. - After Bob withdraws the funds, Carol also knows the secret R and uses it to unlock the HTLC on the Bitcoin blockchain to withdraw 10,000 satoshis. ## Splicing Channel splicing is an on-chain rebalancing method where a node closes a channel and then opens a new channel in a single transaction, thereby changing the balance locked in the channel. When a node locks in more funds through this process, it is referred to as "splice in"; if it reduces the locked funds, it is called "splice out." References - https://www.lightspark.com/glossary/splicing ## Rebalancing See SoK: Payment Channel Networks | Protocol | BC | Trusted entity | Graph compatibility | Privacy | Year | | ------------------------- | ----------- | ----------------- | ------------------- | ------- | ---- | | Revive \[103] | TC | Central Trust | Cycles only | No | 2017 | | Subramanian et al. \[104] | TC and UTXO | No | Agnostic | Yes | 2019 | | Rebal \[105] | UTXO | No | Cycles only | Yes | 2021 | | Hide & Seek \[106] | UTXO | Distributed Trust | Cycles only | Yes | 2022 | | Cycle \[107] | TC | No | Cycles only | Yes | 2022 | | Shaduf \[108] | TC | No | Agnostic | Yes | 2022 | | Musketeer \[109] | UTXO | No | Cycles | No | 2023 | | Chen et al. \[110] | UTXO | No | Cycles | No | 2024 | BC: - TC: Turing-complete - UTXO: Bitcoin-like scripting References - \[103] R. Khalil and A. Gervais, “Revive: Rebalancing off-blockchain pay-ment networks,” ser. CCS ’17. Association for Computing Machinery, 2017, p. 439–453. - \[104] L. M. Subramanian, G. Eswaraiah, and R. Vishwanathan, “Rebalancing in acyclic payment networks,” in 2019 17th International Conference on Privacy, Security and Trust (PST). IEEE, 2019, pp. 1–5. - \[105] N. Awathare, V. J. Ribeiro, U. Bellur et al., “Rebal: channel balancing for payment channel networks,” in 2021 29th International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommu-nication Systems (MASCOTS). IEEE, 2021, pp. 1–8. - \[106] Z. Avarikioti, K. Pietrzak, I. Salem, S. Schmid, S. Tiwari, and M. Yeo, “Hide & seek: Privacy-preserving rebalancing on payment channel networks,” in International Conference on Financial Cryptography and Data Security. Springer, 2022, pp. 358–373. - \[107] Z. Hong, S. Guo, R. Zhang, P. Li, Y. Zhan, and W. Chen, “Cycle: Sustainable off-chain payment channel network with asynchronous re-balancing,” in 2022 52nd Annual IEEE/IFIP International Conference on Dependable Systems and Networks (DSN), 2022, pp. 41–53. - \[108] Z. Ge, Y. Zhang, Y. Long, and D. Gu, “Shaduf: Non-cycle and privacy-preserving payment channel rebalancing,” Cryptology ePrint Archive, 2022. - \[109] Z. Avarikioti, S. Schmid, and S. Tiwari, “Musketeer: Incentive-compatible rebalancing for payment channel networks,” Cryptology ePrint Archive, 2023. - \[110] W. Chen, X. Qiu, Z. Cai, B. Tang, L. Du, and Z. Zheng, “Graph neural network-enhanced reinforcement learning for payment channel rebalancing,” IEEE Transactions on Mobile Computing, vol. 23, no. 6, pp. 7066–7083, 2024. ### AI Delving Into #### Key Takeaways The landscape of off‐chain payment channel rebalancing has evolved rapidly since 2022, moving beyond simple cyclic rebalancing to incorporate multi‐party, split, star‐shaped, and learning‐based strategies. Recent work emphasizes: - **Non‐cycle and multi‐channel approaches** (Split, Shaduf++, Starfish) that reduce dependence on cyclic topologies while enhancing privacy and efficiency. - **Incentive alignment** (Musketeer) ensuring participants profit from rebalancing operations. - **Machine learning integration** (DRL-PCR, GNN-RL) to adaptively optimize rebalancing decisions under dynamic network conditions. #### Non-Cycle and Split Multi-Channel Rebalancing **Shaduf++** (2025) extends Shaduf by allowing direct fund transfers among arbitrary channels without requiring cycles. It preserves privacy by hiding cycle information and significantly reduces user deposits while maintaining throughput​​.[1] **Split** (2025) introduces a *split multi-channel* strategy that divides a rebalancing amount across several channels selected based on traffic load weights and minimal loss probabilities. By distributing pressure, Split achieves higher success rates and better balance restoration compared to Revive and Shaduf​​.[2] #### Star-Shaped and Incentive-Compatible Protocols **Starfish** (Apr. 2025) leverages a *star topology* to concentrate rebalancing on high-capacity hub nodes. This shape maximizes rebalancing efficiency by minimizing path lengths and liquidity lock-up durations across multiple channels​​.[3] **Musketeer** (2023) formulates rebalancing as a game‐theoretic mechanism, designing fees and reward schemes that ensure all participants have positive utility, thus aligning incentives for off-chain fund redistribution​​. #### Learning-Based Rebalancing **DRL-PCR** (2024) combines *deep reinforcement learning* with graph neural networks to predict optimal rebalancing actions, avoiding local optima and adapting to network state changes. The approach outperforms greedy baselines on simulated Lightning snapshots​​.[4] **Deep Scheduling Strategy** (2025) uses reinforcement learning to schedule rebalancing tasks by forecasting transaction demands and channel capacities. It shows improved throughput and reduced channel depletion events compared to static heuristics​​.[5] **Hybrid Pathfinding with RL** (2025) integrates deterministic pathfinding with an RL agent to select routes that balance success probability and liquidity usage. Experimental results demonstrate superior payment success under stochastic failure distributions​​.[6] #### Admission Control and Cost-Aware Rebalancing **R2** (2023) addresses off-chain rebalancing from an intermediary’s viewpoint, developing online algorithms for admission control, on-chain recharging, and off-chain rebalancing with linear fee models. The proposed method minimizes overall routing and rebalancing costs under realistic demand patterns​​.[7] #### Conclusion The field is shifting toward flexible, privacy-preserving rebalancing schemes that exploit multi-party coordination, star-shaped topologies, and machine learning. Future work will likely focus on reducing computational overhead, integrating real‐world network dynamics, and deploying incentive-aligned protocols in production Lightning nodes. #### References - [1](https://www.computer.org/csdl/journal/tq/2025/02/10606068/1YMKfR8YlNe) - [2](http://ijns.jalaxy.com.tw/contents/ijns-v27-n4/ijns-2025-v27-n4-p838-849.pdf) - [3](https://arxiv.org/abs/2504.20536) - [4](https://www.semanticscholar.org/paper/d0541396d4586347a852b3cc113c10977abd6e7a) - [5](https://ui.adsabs.harvard.edu/abs/2025ITNet..33..570R/abstract) - [6](https://www.sciencedirect.com/science/article/abs/pii/S0952197625002258) - [7](https://fc23.ifca.ai/preproceedings/57.pdf) - [8](https://www.ndss-symposium.org/ndss-paper/auto-draft-254/) - [9](https://www.lightspark.com/glossary/channel-rebalancing) - [10](https://cryptoslate.com/why-lightning-network-capacity-declining-20-in-2025-is-not-as-bad-as-it-sounds/) - [11](https://scholar.google.co.in/citations?user=jLr_xi4AAAAJ&hl=vi) - [12](https://dl.acm.org/doi/10.1145/3133956.3134033) - [13](https://www.tryspeed.com/blog/whats-next-for-usdt-on-lightning/) - [14](https://www.semanticscholar.org/paper/Revive:-Rebalancing-Off-Blockchain-Payment-Networks-Khalil-Gervais/7a2918f9f0192e9a83c46c1ee58742dd6bd98b87) - [15](https://www.ndss-symposium.org/wp-content/uploads/2025-495-paper.pdf) - [16](https://www.fidelitydigitalassets.com/sites/g/files/djuvja3256/files/acquiadam/FDA_TheLightningNetwork_ExpandingBitcoinUseCases_1187503.1.0_V5.pdf) - [17](https://dl.acm.org/doi/10.1145/3721146.3721965) - [18](https://aurpay.net/aurspace/lightning-network-enterprise-adoption-2025/) - [19](https://www.londonstockexchange.com/news-article/PR1/1-million-placing-to-expand-lightning-network/17155854) - [20](https://onekey.so/blog/ecosystem/what-is-the-lightning-network/) ## Node Selection To maximize the reach and routing success as a new Lightning user, it's recommended to open channels with well-connected, highly reliable routing nodes. - Community picked - Analyze the network and generate a leaderboard ## L1/L2 Swap Alice pays Bob via L1, Bob pays back Alice via Fiber. ## Further Readings - https://www.perplexity.ai/search/what-s-the-latest-trends-of-th-ajv1YFtBT32QzUgAtO_IUg - Horcrux - S. S. Sahoo, M. M. Hosmane, and V. K. Chaurasiya, “A secure payment channel rebalancing model for layer-2 blockchain,” Internet of Things, vol. 22, p. 100822, 2023. - P. Li, T. Miyazaki, and W. Zhou, “Secure balance planning of off-blockchain payment channel networks,” in IEEE INFOCOM 2020-IEEE conference on computer communications. IEEE, 2020, pp. 1728–1737.