# LTRAS: A Linkable Threshold Ring Adaptor Signature Scheme for Efficient and Private Cross-Chain Transactions

Crypto.world

> Adaptor signatures let two parties exchange signatures that only become valid once a secret is revealed, which makes them a natural fit for atomic cross-chain swaps without hash-timelock contracts.

**arXiv**: 2602.05431

**Published**: 2026-02-05

**Authors**: Yi Liang, Jinguang Han

## AI summary

LTRAS: A Linkable Threshold Ring Adaptor Signature Scheme

## Overview

**Research area:** Applied cryptography for blockchains — specifically adaptor signatures, ring/threshold ring signatures, and cross-chain atomic swap protocols.

**Technical level:** Advanced. The paper is written for readers comfortable with discrete-logarithm-based cryptography, sigma protocols, security games, and UTXO-based blockchain accounting models. The definitions and security experiments (Algorithms 1–3) are formal.

**Scope (one sentence):** The paper defines, constructs, proves secure, and evaluates a linkable threshold ring adaptor signature scheme that hides multiple signing accounts inside one ring while retaining the conditional-binding and double-spending-detection properties needed for private cross-chain atomic swaps.

Note on sourcing: the supplied text is truncated partway through the security-model section, so Section 5 (Performance Evaluation) appears only through the abstract and introduction's qualitative claims. No concrete running times, signature byte sizes, curve choices, implementation languages, or benchmark hardware appear in the available content, and none are supplied below.

## What This Paper Is About

Adaptor signatures let two parties exchange signatures that only become valid once a secret is revealed, which makes them a natural fit for atomic cross-chain swaps without hash-timelock contracts. The problem is that a standard adaptor signature exposes the signer's public key and account address on-chain, so observers can build transaction graphs and de-anonymize users. Existing ring adaptor signature schemes solve part of this but hide only a single account per ring, so a user in a UTXO-based system like Monero — who typically holds many one-time public keys and spends several at once — must build one ring per account, which blows up signature size and communication cost. The paper's goal is a single primitive that combines adaptor-signature conditional binding, threshold ring signatures for multi-account payments, and linkability for double-spending detection.

## Key Contributions

1. **A new primitive: LTRAS.** A linkable ((t,n))-threshold ring adaptor signature scheme, defined over a hard relation and a standard signature scheme, with the algorithm suite Setup, KeyGen, GenR, PreSign, PreVerify, Adapt, Verify, Ext, and Link. Uniquely, GenR outputs a *pair* of statements (\boldsymbol{W} = (W_1, W_2)) sharing one common witness (w), and Ext succeeds only if both ((W_1, w) \in \mathcal{R}) and ((W_2, w) \in \mathcal{R}).
2. **Four properties in one construction.** Adabtability and extractability (from adaptor signatures), anonymity (from ring signatures), joint payment support (from threshold ring signatures, hiding multiple public keys in the same ring and completing a transfer with one signing operation), and linkability (detecting signatures made by the same private key to prevent double-spending). Table 1 shows no prior scheme in the compared set holds all five properties — adaptability, witness extractability, anonymity, linkability, and joint payments — simultaneously.
3. **Formal security treatment.** A formal definition, correctness and pre-signature adaptability definitions, and three security experiments: `aWitExt` (witness extractability, a four-game model where the adversary also supplies the statement), `aSignForge` (unforgeability under a threshold corruption bound (|\boldsymbol{PK}^* \cap \boldsymbol{F}| \le t-1)), and `aAnon` (anonymity, a five-game model). Security proofs accompany them.
4. **Performance analysis plus a cross-chain application discussion.** Both theoretical analysis and an experimental implementation are reported, along with a discussion of use in cross-chain atomic swaps, extending compatibility to ring-signature blockchains such as Monero.

## Main Findings

- **Dominance over prior schemes on combined properties:** According to Table 1, ([1]) (Aumayr et al.) and ([27]) (functional adaptor signatures) lack anonymity, linkability and joint payments; ([22]) (Qin et al., the first linkable ring adaptor signature) and ([28]) (Wang et al., generic linkable ring adaptor signature) add anonymity and linkability but not joint payments; ([25]) (Sui et al.) has anonymity but neither linkability nor joint payments; and ([8]) (Duan et al.) and ([34]) (Zheng et al.) have anonymity, linkability and joint payments but neither adaptability nor witness extractability. Only LTRAS holds all five.
- **The single-account bottleneck is the gap being closed:** The stated limitation of ([22]) is that it can hide only a single account in a ring and its signature size grows significantly with the number of payment accounts, making it impractical for multi-account use; ([28])'s generic construction is likewise described as unsuitable for multi-account scenarios.
- **Claimed overhead reduction:** The abstract and introduction state that experiments show "significantly lower computation and communication overhead compared to existing schemes in large ring sizes and multi-account payment scenarios." The specific measured values, baselines, ring sizes and timings are not present in the supplied (truncated) text and are therefore not reported here.
- **Mixing adaptor signatures with ring signatures requires a threshold ring:** Because a Monero-style user often needs several accounts as inputs, standard ring signatures force one ring per account, while threshold ring signatures, many-out-of-many proofs ([7]) and any-out-of-many proofs ([34]) can hide multiple accounts in a single ring and reduce signature size.
- **Rogue-key resistance is built into the transformation:** The scheme adopts the sliding window transformation from Duan et al. ([8]), computing (d = H(\boldsymbol{PK})) and (y_i = \prod_{k=i \bmod n}^{(i+t-1) \bmod n} pk_k^{,d}), so that the hash digest (d) mitigates rogue-key attacks when aggregating (t) consecutive public keys.
- **Adjacent work does not cover the same ground:** AucSwap ([16]) relies on hash-timelock contracts and lacks privacy-preserving conditional binding; ZeroMT ([24]) uses aggregated zero-knowledge proofs for intra-chain multi-transfer privacy but cannot support conditional asset release; Liu et al. ([17]) is quantum-resistant and oracle-based but confined to oracle-initiated conditional payments with no cross-chain support.

## Methodology in Plain English

The authors start from the standard adaptor signature recipe: a "pre-signature" is produced relative to a hard mathematical statement, and anyone holding the matching secret can turn it into a real signature — at which point an observer can recover that secret by comparing the pre-signature and the final signature. They extend this in three ways. First, they make the hard relation two-headed: key generation outputs two statements sharing one witness, so extracting the witness requires both statements to check out, which is what lets the secret be bound to a whole multi-account payment rather than a single key. Second, they borrow the sliding-window transformation from Duan et al. to convert a single-signer ring into a ((t, n))-threshold ring, so that (t) of (n) listed public keys can jointly sign and a user's several accounts sit in the same ring — this is where the efficiency win comes from. Third, they add a Link algorithm that compares two signatures and reports whether they came from the same signer, without revealing who that signer is. They then write down correctness, adaptability, witness extractability, unforgeability and anonymity as formal games with a challenger and an adversary, prove the construction meets them, and finally implement it to compare computation and communication costs against the prior schemes in Table 1.

## Why This Matters

**Impact on research.** The paper argues that adaptor signatures, despite being a leading primitive for off-chain channels, leave signer identity exposed, and that the ring-signature line of work has evolved separately — linkable ring adaptor signatures ([22, 28]), linkable threshold ring signatures ([8]), any-out-of-many proofs ([34]) — without ever merging threshold rings, linkability, and adaptor conditional binding. LTRAS is positioned as closing that specific gap and gives the community a formal definition and security-model template for the combined primitive.

**Real-world applications (from the paper's framing):**

- **Cross-chain atomic swaps with ring-signature chains** — the paper's Section 6 is dedicated to this, and the scheme's design is explicitly aimed at blockchains like Monero where standard adaptor-signature swaps do not fit.
- **Multi-account joint payments in UTXO systems** — combining several one-time public keys into one ring and signing once, instead of one ring per input account.
- **Off-chain payment channels and payment channel networks** — cited as the primary motivation for adaptor signatures ([28]) and for avoiding complex hash/timelock scripts.
- **Double-spending prevention with retained anonymity** — Link detects two signatures from the same private key without identifying the key's owner.

**Industry relevance.** Because adaptor signatures rely only on digital-signature scripts that blockchains already support, the paper argues they improve compatibility with chains whose scripting languages are limited and reduce on-chain fees compared with hash-timelock-based constructions. That is a direct argument to wallet and exchange developers who need privacy-preserving swaps on script-constrained chains.

## Future Directions

- **Post-quantum LTRAS.** The paper cites post-quantum adaptor signatures ([26, 13]) and a quantum-resistant oracle-based conditional payment scheme ([17]) as existing alternatives; LTRAS itself is built on the discrete logarithm assumption, which is not quantum-resistant, and no lattice- or code-based analogue is presented.
- **Filling in granularity.** Vanjani et al.'s functional adaptor signatures ([27]) extract a *function* of the witness rather than the whole witness, giving finer-grained access control; whether an LTRAS variant can support partial witness release is not addressed.
- **Witness hiding.** Liu et al. ([18]) argued that unconditional witness exposure limits applications and built witness-hiding adaptor signatures — a property LTRAS does not claim and which could conflict with linkability.
- **Concrete deployment evaluation.** The supplied text does not report concrete costs, ring-size sweeps, or baselines, so independent benchmarking and an integration study against a real Monero-style or script-limited chain remain open.

## Target Audience

Cryptography and blockchain-security researchers working on adaptor signatures, ring signatures, or atomic swap protocols, who will find the formal definition and the three security experiments the most reusable part. It also suits protocol designers and wallet engineers on UTXO-based privacy chains who need multi-input payments with adaptor-style atomicity. Readers without a discrete-log and security-game background will find the definitions heavy; practitioners mainly interested in measured performance will need the full paper, since the experimental figures are not in the available excerpt.

## Authors’ abstract

Despite the advantages of decentralization and immutability, blockchain technology faces significant scalability and throughput limitations, which has prompted the exploration of off-chain solutions like payment channels. Adaptor signatures have been considered a promising primitive for constructing such channels due to their support for atomicity, offering an alternative to traditional hash-timelock contracts. However, standard adaptor signatures may reveal signer identity,…

[The full abstract on arXiv](https://arxiv.org/abs/2602.05431)

[Read the original paper](https://arxiv.org/abs/2602.05431)

---

Canonical: https://crypto.world/paper/ltras-a-linkable-threshold-ring-adaptor-signature-scheme-for-efficient-2602.05431
