Paper 2023/116

A Cryptographic Layer for the Interoperability of CBDC and Cryptocurrency Ledgers

Abstract

Cryptocurrencies are used in several, distinct use cases, thereby sustaining the existence of many ledgers that are heterogeneous in terms of design and purpose. In addition, the interest of central banks in deploying Central Bank Digital Currency (CBDC) has spurred a blooming number of conceptually different proposals from central banks and academia. As a result of the diversity of cryptocurrency and CBDC ledgers, interoperability, i.e., the seamless transfer of value between users that operate on different ledgers, has become an interesting research problem. In fact, interoperability has been explored both in CBDC and cryptocurrencies, and numerous proposals exist. However, these proposals are tailored to the characteristics of the ledgers for which they are designed. For instance, some rely on trusted hardware, others rely on the scripting capabilities of the underlying ledger or on specific cryptographic assumptions of the transaction authorization mechanism (e.g., adaptor signatures), while others rely on a trusted entity. This fragmentation results in the repetitive development of interoperablity protocols that address the same applications across the various ledgers. In this work, we propose an alternative approach: decouple the transaction authorization details of each ledger from the definition of cross-ledger applications. To do so, we define a middle layer that abstracts the core functionality for authorizing transactions in any ledger and the security notions of interest. This middle layer serves two purposes: (i) it becomes the main cryptographic building block to (re)define cross-ledger applications in a ledger agnostic manner; (ii) for any ledger that exists (and the new to come), it suffices to prove that it is a secure instance of the middle layer to be compatible with cross-ledger protocols. We define two new primitives for our middle layer, the basic payment ledger (BL) and the conditional payment ledger (CL). We prove that the two most common transaction authorization mechanism (digital signatures and zero knowledge proofs) are secure constructions for BL. We also prove that common smart contracts (e.g., HTLC and PTLC) and the combination of adaptor signatures with verifiable timelock puzzles are also secure constructions for CL. Finally, we discuss how to design some popular applications (e.g. atomic swaps between ledgers) using our middle layer.