Your Trust, Your Terms: A General Paradigm for Near-Instant Cross-Chain Transfer

arXiv:2403.15191 · cs.CR, cs.DC · Submitted 2024-03-22 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Security Radio. Generated commentary on the latest security and cryptography papers.

Nadia: We're looking at the paper: "Your Trust, Your Terms".

Elias: Here's the gist of the paper. Cross-chain transactions today remain slow, costly, and fragmented Neither custodial exchanges nor non-custodial liquidity bridges guarantee users a unilateral path to recover assets if the infrastructure fails The Delegated Ownership Transfer (DOT) paradigm decouples key ownership from value ownership to enable secure, high-performance cross-chain payments This paradigm introduces a sandboxed on-chain Temporary Account (TA) for value ownership while delegating its private key to an abstract Trusted Entity (TE) Security follows from dual guarantees: the TE’s exclusive control prevents double-spending, while a pre-signed, unilateral recovery transaction ensures users retain ultimate authority over their assets Core Paradigm and Design Goals The DOT paradigm is built on the core technical insight of decoupling key ownership from value ownership A payment is reframed as a near-instant, off-chain ownership handoff of the TA The architecture provides two synergistic properties for the TA: Sandboxed Control, where the control key is held exclusively by the TE, resolving doublespending, and Guaranteed Liveness via Unilateral Recovery, which mandates a recovery transaction signed by the TE to ensure asset recoverability This separation achieves the four essential properties of security, performance, universality, and user-defined trust Instantiation Architectures The paradigm offers two concrete instantiations to realize user-defined trust The first is the TEE-based TE, anchoring trust in hardware security and remote attestation This instantiation avoids direct TEE↔TEE networking by having users relay all messages, which reduces latency and dependencies The second is the DTC-based TE, anchoring trust in cryptographic assumptions via a distributed group of nodes This instantiation relies on a threshold set to t = ⌈2n/3⌉, guaranteeing correctness and liveness against up to f < n/3 malicious nodes Security Formalization The security of the DOT paradigm is formalized within the Universal Composability framework (UC) The protocols ΠTEE and ΠDTC GUC-realize the ideal functionality FDOT in their respective hybrid models The simulator S can produce an interaction indistinguishable from the real protocol for any adversary A, capturing the nature of faulty TE and asynchronous networks Theorem 1 proves that these protocols realize FDOT under the assumption of unforgeability of the signature scheme and security of underlying primitives Performance Results The geo-distributed prototype shows that cross-chain payments complete in under 16.70 ms and atomic swaps in under 33.09 ms These results demonstrate that costs are fully decoupled from Layer-1 gas fees In a test of 1,000 sequential transactions, DOT-T achieved an average latency of just 33.

Title and authors: Nadia: So we're looking at this paper right now titled Your Trust, Your Terms: A General Paradigm for Near-Instant Cross-Chain Transfer. It’s addressing the slow, costly, and fragmented state of cross-chain transactions we talked about before.

Elias: Right. The authors are tackling the problem where neither custodial exchanges nor non-custodial liquidity bridges give users a guaranteed way to get their assets back if the whole infrastructure goes down. They introduce this Delegated Ownership Transfer or DOT paradigm to solve that fundamental weakness.

Priya: It sounds like they’re proposing a way to separate who owns the key from who actually owns the value on chain, which is a really interesting starting point for security.

Nadia: Exactly. The core idea is decoupling key ownership from value ownership so we can get high-performance cross-chain payments without having to rely on slow Layer one gas fees for every single transfer.

Elias: They achieve this by setting up a sandboxed on chain temporary account, which holds the value, but they delegate the private key for that account to some abstract Trusted Entity or TE.

Priya: So if I understand this right, the value is on the ledger in this temporary account, but I don't have direct control of it because someone else—this TE—holds the key?

Nadia: Precisely. The paper calls this a near-instant off chain ownership handoff of that temporary account. They frame payments and swaps as happening quickly off chain before settling the actual ownership transfer on the blockchain.

Elias: And they build in two properties for that temporary account: sandboxed control, where the TE holds the control key to stop double spending, and guaranteed liveness through a unilateral recovery transaction signed by the TE.

Priya: That sounds like they are trying to solve both sides of the coin: you get speed but you still have a safety net if things go wrong or if the system gets stuck.

Nadia: It’s about separating verification from global consensus so we can get network-speed payments, but they say that’s not enough if the payer's key still directly controls the live balance.

Elias: They argue that separating key control from the on chain value without an accountable verification path just leaves you with race conditions or slow finality checkpoints, which is a problem.

Title and authors: Priya: I wonder how this affects someone who just wants to move assets quickly across different systems without having to wait for a global consensus mechanism every time.

Nadia: It means the latency and fees are decoupled from Layer one gas fees, and their geo distributed prototype shows cross chain payments completing in under sixteen point seventy milliseconds.

Elias: And they also showed atomic swaps can complete in under thirty-three point nine milliseconds, which is pretty fast for what it achieves.

Priya: Those numbers are impressive, especially when you consider how fragmented cross-chain stuff currently is and how slow traditional bridges have been.

Nadia: They also show that in a test of one thousand sequential transactions, DOT T achieved an average latency of just thirty-three milliseconds <ref:2403.15191#pg1>. That’s really low for this kind of operation.

Elias: The authors use two instantiations to realize user defined trust: one based on a TEE called Trusted Execution Environment, and another based on a DTC called Distributed Threshold Cryptography.

Priya: So they are offering two different ways for users to anchor their trust in this system, depending on whether they prefer hardware security or decentralized cryptography.

Nadia: That’s right. The TEE-based TE anchors trust in hardware security and remote attestation, avoiding direct TEE to TEE networking by having users relay all messages which reduces latency and dependencies.

Elias: And the DTC-based TE relies on cryptographic assumptions via a distributed group of nodes with a threshold set to ceiling of two over n divided by three, guaranteeing correctness against up to f less than n over three malicious nodes.

Priya: That sounds like they’re trying to balance that speed and security trade off between those two models.

Nadia: From a security standpoint, they formalized the DOT paradigm using the Universal Composability framework, showing that protocols ΠTEE and ΠDTC realize the ideal functionality FDOT in their respective hybrid models.

Elias: They proved this under the assumption of unforgeability of the signature scheme and security of underlying primitives, meaning they’ve done a solid job on the theoretical side.

Priya: But what about how secure it is when you look at the actual data flow? What are some of those specific security concerns they addressed in that formalization?

Title and authors: Nadia: They address how the TE’s exclusive control prevents double spending, which is one big thing, and a pre signed unilateral recovery transaction ensures users still have ultimate authority over their assets.

Elias: That unilateral recovery path is what keeps them from being just another fast but fragile system; it mandates a recovery transaction signed by the TE to ensure asset recoverability.

Priya: So for someone who only listens to the show, what does this mean practically? Does it mean my funds are safer when I use this DOT setup compared to sending them through a standard bridge right now?

Nadia: It means your funds are protected by that dual guarantee: the TE’s control and that unilateral recovery option. It’s about retaining ultimate authority over your assets even if the main infrastructure has issues.

Elias: They also discussed how they can enforce delayed recovery on any ledger by employing Time Lock Puzzles, which is a blockchain agnostic primitive allowing users to decrypt and broadcast after the timeout without relying on chain specific features.

Priya: That time lock idea is interesting because it means the system isn't locked into one specific blockchain’s time mechanism for recovery purposes.

Nadia: It’s about adaptability, letting you enforce delayed recovery on any ledger by transforming the recovery mechanism into a call to a dedicated smart contract that enforces the time lock logic if you are using an EVM chain.

Elias: They also showed how they can bridge these different trust models, moving from a TEE based TE to a DTC based one using Verifiable Secret Sharing schemes, enabling extensibility to heterogeneous interoperability.

Priya: Extensibility sounds good for the future, but I wonder if managing those different trust anchors adds complexity that might introduce new kinds of vulnerabilities somewhere else.

Nadia: That's a fair point about complexity, but they claim it achieves the four essential properties of security, performance, universality, and user defined trust through this separation.

Elias: And economically speaking, they use Taproot based batch operations for both account setup and liquidation to dramatically amortize on chain costs through symmetrical batching principles.

Priya: So the economic efficiency comes from using those batch operations to make the on chain costs much smaller over time, which is a key practical concern for adoption.

Title and authors: Nadia: It’s about making it economically viable by amortizing those on chain costs through symmetrical batching principles when setting up or liquidating accounts.

Elias: So to wrap this up, the DOT paradigm introduces a sandboxed temporary account for value ownership and delegates the key to a TE, securing it with exclusive control and a unilateral recovery path.

Priya: It seems like they are providing a more robust security posture than what’s currently available in terms of cross-chain asset recovery.

Nadia: Absolutely. The DOT paradigm offers secure high performance cross chain payments completing in under sixteen point seventy milliseconds, with costs decoupled from Layer one gas fees by reframing the payment as an off chain transfer of temporary account ownership.

Elias: And they showed robust atomic swaps with guaranteed recovery paths that remain fair even if the Trusted Entity fails, ensuring users retain ultimate authority over their assets via a pre signed unilateral recovery transaction.

Priya: The flexible trust anchors using both TEE and DTC models give users a choice in how they want to anchor their trust, from hardware security to decentralized cryptography.

Nadia: They also offer enhanced security in TEE based systems by protecting the temporary account secret key inside an isolated enclave, which guarantees the confidentiality and integrity of code and data within it even from a privileged OS.

Elias: And with distributed threshold cryptography, they can operate without a trusted dealer where any subset of size at least t can perform authorized operations while the secret key is never reconstructed in the clear.

Priya: They also introduced blockchain agnostic time lock mechanisms using Time Lock Puzzles as a way to enforce delayed recovery on any ledger by decrypting and broadcasting only after the timeout.

Nadia: And they adapt on chain settlement for EVM chains by transforming the recovery mechanism into a call to a dedicated smart contract that enforces the time lock logic, making DOT adaptable to programmable off chain transfers.

Elias: Finally, they achieve economic efficiency through batching operations using Taproot based methods to dramatically amortize on chain costs through symmetrical batching principles for account setup and liquidation.

Priya: So this paper provides a comprehensive framework that tackles the core issues of speed, security, and recovery in cross-chain transactions by separating key control from value ownership.

The paper's summary: Nadia: So we’re looking at the summary of this DOT paper now, which boils down to decoupling who owns the key from who actually owns the value on chain so we can get these cross-chain transfers under sixteen point seventy milliseconds, and costs are separated from Layer one gas fees.

Elias: Right. They frame a payment not as moving money directly on the ledger, but as a near-instant off chain handoff of that temporary account ownership. It’s about getting the speed by handling the ownership transfer off chain first before settling it on the blockchain later.

Priya: From what I’m seeing in this summary, they are building in two safety nets for that value owner: sandboxed control where the trusted entity holds that key to stop double spending, and a guaranteed liveness through a unilateral recovery transaction signed by that same entity.

Nadia: Exactly. It’s not just about speed; it’s about having a clear path back to your assets if the infrastructure gets stuck or if something goes wrong on the ledger side. That means you still have ultimate authority over what you sent even with this separation of duties.

Elias: The authors use two different ways to set up that trusted entity, one based on hardware security, called TEEs, and another based on cryptography from a distributed group of nodes using threshold cryptography. It’s giving users a choice between anchoring trust in physical hardware or in decentralized math.

Priya: I’m interested in the security part of that. They formalize this entire thing using the Universal Composability framework, which means they proved that their protocols actually work under a lot of adversarial conditions, even when dealing with faulty entities or networks that aren't perfectly synchronized.

Nadia: That’s what makes it solid; they didn't just show it works in a lab. They used this framework to prove the ideal functionality FDOT is achieved in both their TEE and DTC models, assuming the signature scheme is unforgeable.

Elias: And those performance numbers are what really grab my attention for the engineering side. The geo distributed prototype showed cross-chain payments completing under sixteen point seventy milliseconds, and atomic swaps under thirty-three point zero nine milliseconds. That’s a measurable improvement over current solutions.

Priya: But I gotta ask, how does this affect the actual data we look at? Does that fast settlement actually translate to a more robust system for privacy researchers? Are there any lingering concerns about what the temporary account itself holds while it’s in transit off chain?

Nadia: The summary implies the value ownership is sandboxed on chain within that temporary account, but its control key stays with the trusted entity. The security comes from that dual guarantee: exclusive control from the entity and that unilateral recovery path for you, the user.

Elias: They also mentioned they can use time lock puzzles to enforce delayed recovery on any ledger, which is a blockchain agnostic way to make sure you can’t just instantly recover something if things are messy. It lets you wait out a timeout before broadcasting the final transaction.

Priya: That adaptability with the time lock is interesting because it means they aren't tied to one specific chain’s recovery mechanics, which is important when we're looking at interoperability across many different systems.

Nadia: So, what this means for us listening right now is that if you’re dealing with slow cross-chain transfers or worried about losing access if a bridge fails, this DOT approach offers a path to near-instant settlement with built-in recovery mechanisms.

Elias: It really pushes the idea that we can separate the high performance of the network from the absolute control of our assets in a way that’s mathematically sound.

The paper's improvements: Tom: So we’re looking at how the authors suggest they can make this whole thing even better, moving beyond just the initial paradigm setup.

Nadia: They’re suggesting ways to plug in more different trust models, like adding support for TEEs and also using distributed threshold cryptography for the trusted entity. It gives users a choice on how they want to anchor their security.

Elias: That’s a big deal because it means you can use hardware security if you trust your specific chip, or you can use that decentralized key management approach where any subset of nodes can do authorized operations without reconstructing the secret key in the clear.

Priya: I wonder if adding more anchors makes things more complex to audit? When we look at data, we want to make sure every layer of trust is transparent enough for us to verify it properly.

Nadia: They address that by showing how they can bridge those different trust models using verifiable secret sharing schemes between the TEE-based entity and the DTC-based one. It’s about extensibility to heterogeneous interoperability, letting you connect different types of systems together.

Elias: And for the EVM chains, they acknowledge that some of those chains don't have native time locks for recovery, so they suggest adapting the recovery mechanism into a call to a dedicated smart contract that enforces the time lock logic. That makes it adaptable to programmable off chain transfers.

Priya: It’s interesting because it shows how you can make this system flexible enough to work on different types of ledgers, not just one specific type of blockchain infrastructure.

Nadia: And they also tackle the economic side by using Taproot based batch operations for both account setup and liquidation, which is meant to dramatically amortize those on chain costs through symmetrical batching principles. That’s about making it economically viable over time.

Elias: So, the authors are trying to build a system that's not just fast and secure in one way, but flexible enough to adapt its trust anchor and its settlement mechanism depending on what kind of blockchain you’re using or what security level you need.

Priya: It sounds like they’re aiming for something that can handle the messy reality of real-world cross-chain data flows, not just a theoretical setup.

Nadia: Exactly. This moves it from a single solution to a more general framework that lets different people tailor the security and performance based on their specific needs and where they are operating.

Conclusion: Nadia: So we’re wrapping up this talk on "Your Trust, Your Terms: A General Paradigm for Near-Instant Cross-Chain Transfer" by summarizing what they actually achieved and where this leaves us.

Elias: It boils down to using a delegated ownership transfer model to decouple key control from value ownership, giving you near-instant cross-chain payments under sixteen point seventy milliseconds, which is a big speed bump compared to current solutions.

Priya: I think the real implication for us is seeing how they build in that dual guarantee: the trusted entity’s exclusive control and that unilateral recovery transaction so users keep ultimate authority even if infrastructure fails.

Nadia: That's right. It’s not just about getting fast; it’s about making sure your assets are recoverable, which addresses a major weakness in current cross-chain bridges where things can get stuck forever.

Elias: They offered two ways to anchor that trust, so you can choose between hardware security with TEEs or cryptography using distributed threshold cryptography for managing the trusted entity. It shows flexibility in how you build that trust layer.

Priya: From a data perspective, the fact that they proved this works under Universal Composability means we have a solid theoretical foundation to trust their results about functionality, even if we still need to look closely at the actual data flows for privacy checks.

Nadia: Exactly. We’re seeing how they use things like time lock puzzles and batching operations to make it economically viable and adaptable across different blockchains, which is where the practical stuff comes in.

Elias: So, "Your Trust, Your Terms" gives us a clear blueprint for separating performance from ownership control in these complex systems. It’s a lot of heavy lifting done on the protocol side.

Priya: It changes how we think about asset recovery; it moves it from being a single point of failure to something that has built-in redundancy and user-defined fallback mechanisms.

Nadia: And for those of you listening, this means that as we build cross-chain applications, we have a more robust toolset now to ensure speed doesn't come at the cost of losing control over what you’re moving.

Elias: We’ve got some interesting stuff coming up next, so keep your ears tuned for our discussion on how AI is actually being used to create and defend against malicious code in software.

Priya: We’ll be talking about how these LLMs can be jailbroken into generating harmful code, so we look at the security risks there.

Di Wu, Jingyu Liu, Xuechao Wang, Jian Liu, Yingjie Xue, Kui Ren, Chun Chen

Zhejiang University · Hong Kong University of Science and Technology (Guangzhou) · University of Science and Technology of China

cs.CR, cs.DC

Submitted: 2024-03-22

Updated: 2026-11-03

Comments: arXiv admin note: This submission has been withdrawn because it does not meet arXiv's research content quality standards

Code: https://github.com/ethereum/wiki

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 100/100

The gist: Cross-chain transactions today remain slow, costly, and fragmented Neither custodial exchanges nor non-custodial liquidity bridges guarantee users a unilateral path to recover assets if the…

Key concepts

Delegated Ownership Transfer (DOT) Paradigm
A system that separates who controls the key from who owns the value. A payment is treated as an off-chain handoff of a Temporary Account's ownership. This allows for near-instant transfers because the actual movement of assets happens off-chain, while security is maintained through delegated control to a Trusted Entity (TE).
Temporary Account (TA)
A sandboxed on-chain account used specifically for holding value during cross-chain transfers. Its private key is not held by the user but is delegated to the TE. This separation isolates the value from direct user control, allowing for secure management and delegation of funds across different blockchains without exposing full ownership keys.
Trusted Entity (TE)
An abstract entity that holds the private key controlling a Temporary Account. The TE's role is twofold: it exclusively controls the TA to prevent double-spending, and it must sign a unilateral recovery transaction if needed. This delegation ensures performance while maintaining ultimate user authority through a defined recovery mechanism.

Terminology

Summary

Cross-chain transactions today remain slow, costly, and fragmented Neither custodial exchanges nor non-custodial liquidity bridges guarantee users a unilateral path to recover assets if the infrastructure fails The Delegated Ownership Transfer (DOT) paradigm decouples key ownership from value ownership to enable secure, high-performance cross-chain payments This paradigm introduces a sandboxed on-chain Temporary Account (TA) for value ownership while delegating its private key to an abstract Trusted Entity (TE) Security follows from dual guarantees: the TE’s exclusive control prevents double-spending, while a pre-signed, unilateral recovery transaction ensures users retain ultimate authority over their assets

Core Paradigm and Design Goals

The DOT paradigm is built on the core technical insight of decoupling key ownership from value ownership A payment is reframed as a near-instant, off-chain ownership handoff of the TA The architecture provides two synergistic properties for the TA: Sandboxed Control, where the control key is held exclusively by the TE, resolving doublespending, and Guaranteed Liveness via Unilateral Recovery, which mandates a recovery transaction signed by the TE to ensure asset recoverability This separation achieves the four essential properties of security, performance, universality, and user-defined trust

Instantiation Architectures

The paradigm offers two concrete instantiations to realize user-defined trust The first is the TEE-based TE, anchoring trust in hardware security and remote attestation This instantiation avoids direct TEE↔TEE networking by having users relay all messages, which reduces latency and dependencies The second is the DTC-based TE, anchoring trust in cryptographic assumptions via a distributed group of nodes This instantiation relies on a threshold set to t = ⌈2n/3⌉, guaranteeing correctness and liveness against up to f < n/3 malicious nodes

Security Formalization

The security of the DOT paradigm is formalized within the Universal Composability framework (UC) The protocols ΠTEE and ΠDTC GUC-realize the ideal functionality FDOT in their respective hybrid models The simulator S can produce an interaction indistinguishable from the real protocol for any adversary A, capturing the nature of faulty TE and asynchronous networks Theorem 1 proves that these protocols realize FDOT under the assumption of unforgeability of the signature scheme and security of underlying primitives

Performance Results

The geo-distributed prototype shows that cross-chain payments complete in under 16.70 ms and atomic swaps in under 33. p. 009 ms These results demonstrate that costs are fully decoupled from Layer-1 gas fees In a test of 1,000 sequential transactions, DOT-T achieved an average latency of just 33. p. 0

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Secure High-Performance Cross-Chain Payments via Delegated Ownership Transfer (DOT)

The system can execute cross-chain payments in under 16.70 ms, achieving sub-second perceived settlement with costs decoupled from Layer-1 gas fees by reframing a payment as a near-instant off-chain transfer of Temporary Account ownership.

  1. Robust Atomic Swaps with Guaranteed Recovery Paths

The novel off-chain atomic swap protocol remains fair even if the Trusted Entity (TE) fails, ensuring that a pre-signed, unilateral recovery transaction ensures users retain ultimate authority over their assets by embedding dispute resolution into the recovery path.

  1. Flexible Trust Anchors via Dual Instantiations

The AI system can be deployed using two distinct trust models: a high-performance design based on Trusted Execution Environments (TEEs) for hardware security and a cryptographically robust variant leveraging threshold cryptography for decentralized key management.

  1. Enhanced Security in TEE-Based Systems

By leveraging Trusted Execution Environments, the system can protect the secret key of the Temporary Account inside an isolated enclave, ensuring isolated execution, which guarantees the confidentiality and integrity of code and data within it (an enclave) even from a privileged OS.

  1. Decentralized Key Management via Threshold Cryptography

The AI system can operate using Distributed Threshold Cryptography to generate keys and signatures without a trusted dealer, where any subset of size at least t can perform authorized operations while the secret key is never reconstructed in the clear.

  1. Blockchain-Agnostic Time-Lock Mechanisms

The system can enforce delayed recovery on any ledger by employing Time-Lock Puzzles (TLPs) as a blockchain-agnostic primitive, allowing users to decrypt and broadcast the transaction only after the timeout, without relying on chain-specific features.

  1. Universal Interoperability via Hybrid Trust Bridging

The AI system can bridge disparate trust models by realizing cross-chain transfers from a TEE-based TE to a DTC-based one using Verifiable Secret Sharing (VSS) schemes, enabling extensibility to heterogeneous interoperability.

  1. Adaptable On-Chain Settlement for EVM Chains

For Ethereum Virtual Machine (EVM) chains lacking native time-locked transactions, the system can adapt by transforming the recovery mechanism into a call to a dedicated smart contract that enforces the time-lock logic, making DOT adaptable to programmable off-chain transfers.

  1. Economic Efficiency through Batching

The AI system achieves economic viability by utilizing Taproot-based batch operations for both account setup and liquidation, which allows for dramatically amortize on-chain costs through symmetrical batching principles.

Related papers