OTS-PC: OTS-based Payment Channels for the Lightning Network

Sergio Demian Lerner, Ariel Futoransky · 2025

Read OTS-PC: OTS-based Payment Channels for the Lightning Network by Sergio Demian Lerner, Ariel Futoransky (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-04021.txt.

Introduction

Sergio Demian Lerner and Ariel Futoransky present OTS-PC, a bidirectional payment channel design for the Lightning Network based on one-time signatures over state sequence numbers. Payment channels let two parties transact off-chain, and the widely used Poon-Dryja construction carries well-known complexity in stored data and privacy.

The new construction is described as simpler than Poon-Dryja while keeping compatibility with Lightning routing. Storage per channel stays constant rather than growing with use, and information leakage is kept minimal. Those properties matter for scaling everyday Bitcoin payments, where channels may update many times and operators cannot retain ever-growing revocation data for every channel they hold.

Catalog record

CREATORSergio Demian Lerner, Ariel Futoransky
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE43 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-04021

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

OTS-PC: OTS-based Payment Channels for the Lightning Network Sergio Demian Lerner, Ariel Futoransky Abstract We present a new type of bidirectional payment channel based on One-Time Signatures on state sequence numbers. This new construction is simpler than the Poon-Dryja construction, but provides a number of benefits such as O ⁡ ( 1 ) O(1) storage per channel, minimal information leakage, and compatibility with Lightning Network routing. 1 Introduction Since its inception in 2009, Bitcoin [ 8 ] has demonstrated that decentralized consensus can sustain a global monetary system without trusted intermediaries. However, the protocol’s throughput and latency limitations—roughly seven transactions per second and ten-minute block intervals—pose fundamental constraints on scalability. These parameters are essential to preserve decentralization and security, yet they make Bitcoin unsuitable for high-frequency payments or micro-transactions if used directly on-chain. To overcome these limitations, researchers proposed off-chain payment channels, which allow two parties to conduct multiple transactions without broadcasting each to the blockchain. Only opening and closing the channel require on-chain transactions, while intermediate state updates are exchanged privately and secured by Bitcoin’s scripting primitives. This concept underpins the Lightning Network [ 12 ] , a second-layer protocol that links payment channels to a routed network, allowing instant low-fee payments between participants who do not need to share direct channels. The Lightning Network has proven that layer two constructions can extend Bitcoin’s scalability by several orders of magnitude, while preserving its security guarantees through cryptographic enforcement and dispute mechanisms. However, current payment channel architectures still face limitations: they depend on active monitoring (watchtowers [ 4 , 7 ] ), require prefunded liquidity, and involve complex time-locked contracts that constrain usability and capital efficiency. The notion of payment channels in Bitcoin dates even back to the early versions of the protocol: In the original implementation, Satoshi Nakamoto allowed unconfirmed transactions to be replaced by new versions if all input sequence numbers (nSequence) were increased, allowing multiple interim state changes until a final transaction was broadcast [ 9 ] . Payment channels evolved from the initial ideas, and the Poon-Dryja [ 12 ] design, now standardized in the BOLT specification [ 2 ] , powers the Lightning Network. Improvements continued with new protocols such as Eltoo [ 3 ] , Outpost [ 5 ] or Sleepy [ 1 ] , further reducing the requirements on watchtowers. The latest protocols achieve bidirectional payments, provide HTLCs, infinite-updates and lightweight watchtowers with acceptable privacy. However, Lightning Network channels do not provide the best combination of features: they require peers to store O(log(N)) hashes for N channel updates, but watchtowers still need to store O(N) data. In contrast, Outpost and our proposed protocol require only O(1) storage for peers and watchtowers. Regarding channel functionality, lightning channels do not provide full privacy of the payment channel updates in the presence of watchtowers. From an engineering and practical security point of view, the state of the payment channels is unsatisfactory: the Poon-Dryja design is complex, as is its number of transactions, protocol states, and state transitions. Historically, numerous bugs, incompatibilities with BOLT standards, and security vulnerabilities have been discovered in lightning channel implementations (for example, CVE-2020-26895 in LND). A simpler payment channel design that provides all existing features is enormously valuable. In 2018, Osuntokun presented the idea of Signed Sequence Numbers (SSN) [ 11 ] , a method to reduce the complexity of state revocations and state storage requirements for Lightning Network nodes, using aggregated ECDSA signatures of sequence number commitments embedded in unilateral exit transactions. To the best of our knowledge, Osuntokun’s SSN scheme has not been formalized. In this work, we present OTS Payment Channels (OTS-PCs). OTS-PCs solve the above-mentioned problems and provide the same benefits of SSNs by using hash-based One-Time Signatures. One-time signatures are possible in Bitcoin using the OP_HASH opcode (or any of the other hashing opcodes). Unlike SSNs, our method does not require embedded commitments. The core concept behind OTS-PCs is that each new state has a monotonically increasing sequence number, and the issuer of each state-committing transaction must sign the state sequence number with his OT private key before a payout transaction can be issued in the blockchain to distribute the funds. Once a party signs a state sequence number, the other party is given a window to challenge: if the signed sequence number is not the last, the other party uses a revocation path of the transaction graph to punish the first party by collecting all the funds in the channel. To make OTS-PCs compatible with the Lightning Network, we show how OTS-PCs can support Hash-Time Locked Contracts (HTLC). We demonstrate how we can build lightweight participants and privacy-preserving watchtowers that only need to store a constant amount of information per payment channel. However, to maintain payment privacy, we do not provide a protocol for watchtowers to monitor HTLC timelocks after a successful unilateral exit. OTS-PC channels can be extended to multi-party channels (including channel factories) easily because they have a high amount of symmetry. However, this extension is out of the scope of our work. 2 Specification An OTS-PC is a protocol that allows two parties (Alice and Bob) to manage a payment channel between them. We call Alice and Bob the owners of the channel. The payment channel resembles a shared account where the owners can dynamically re-balance the funds between them. Both owners initially deposit bitcoins in the channel, generally, but not necessarily, in equal amounts. Optional auxiliary parties, W_{a} and W_{b} , act as watchtowers for Alice and Bob, respectively. The payment channel is defined by 3 protocols: setup, state update, and close. An OTS-PC channel state is defined as a funds balance between Alice and Bob at a certain time, and all open HTLC contracts with their properties (more on HTLCs later). Each channel state has a sequence number i , which increments (by one or more) with every channel state update. When the owners agree to move to the next state (e.g., rebalance the shared account), they must co-sign multiple connected transactions. To sign the transactions, the parties can use a 2-out-of-2 multisig or the MuSig2 [ 10 ] scheme. MuSig2 is preferred, as it reduces the on-chain footprint and minimizes information leakage from published transactions. The owners also revoke the previous channel state by creating and signing a set of punishment transactions. The protocol state should not be confused with the channel state. The protocol goes through many protocol state transitions to change the channel state. Protocol participants can perform one of two actions: issue a transaction on the blockchain or send a message directly to another actor (offchain). To send and receive messages offchain, the two owners establish a peer-to-peer authenticated connection. Although the protocol defines which actions each participant can take in every state of the protocol, the publication of a transaction on the blockchain is an asynchronous event. The protocol must consider blockchain events in all states. Therefore, OTS-PC protocol state changes depending on the messages received (which can be informative or just ACKs from a previous message sent) and the transactions issued on the blockchain. Note that the protocol state is not fully determined by onchain transactions. When one or both channel owners want to distribute

Read the complete text