Search for: All records

In distributed systems, a group of learners achieve consensus when, by observing the output of some acceptors, they all arrive at the same value. Consensus is crucial for ordering transactions in failure-tolerant systems. Traditional consensus algorithms are homogeneous in three ways: (1) all learners are treated equally, (2) all acceptors are treated equally, and (3) all failures are treated equally.These assumptions, however, are unsuitable for cross-domain applications, including blockchains, where not all acceptors are equally trustworthy, and not all learners have the same assumptions and priorities. We present the first algorithm to be heterogeneous in all three respects. Learners set their own mixed failure tolerances over differently trusted sets of acceptors. We express these assumptions in a novel Learner Graph, and demonstrate sufficient conditions for consensus. We present Heterogeneous Paxos, an extension of Byzantine Paxos. Heterogeneous Paxos achieves consensus for any viable Learner Graph in best-case three message sends, which is optimal. We present a proof-of-concept implementation and demonstrate how tailoring for heterogeneous scenarios can save resources and reduce latency. 

Securing blockchain smart contracts is difficult, especially when they interact with one another. Existing tools for reasoning about smart contract security are limited in one of two ways: they either cannot analyze cooperative interaction between contracts, or they require all interacting code to be written in a specific language. We propose an approach based on information flow control~(IFC), which supports fine-grained, compositional security policies and rules out dangerous vulnerabilities. However, existing IFC systems provide few guarantees on interaction with legacy contracts and unknown code. We extend existing IFC constructs to support these important functionalities while retaining compositional security guarantees, including reentrancy control. We mix static and dynamic mechanisms to achieve these goals in a flexible manner while minimizing run-time costs. 

The growing adoption of digital assets---including but not limited to cryptocurrencies, tokens, and even identities---calls for secure and robust digital assets custody. A common way to distribute the ownership of a digital asset is (M, N)-threshold access structures. However, traditional access structures leave users with a painful choice. Setting M = N seems attractive as it offers maximum resistance to share compromise, but it also causes maximum brittleness: A single lost share renders the asset permanently frozen, inducing paralysis. Lowering M improves availability, but degrades security. In this paper, we introduce techniques that address this impasse by making general cryptographic access structures dynamic. The core idea is what we call Paralysis Proofs, evidence that players or shares are provably unavailable. Using Paralysis Proofs, we show how to construct a Dynamic Access Structure System (DASS), which can securely and flexibly update target access structures without a trusted third party. We present DASS constructions that combine a trust anchor (a trusted execution environment or smart contract) with a censorship-resistant channel in the form of a blockchain. We offer a formal framework for specifying DASS policies, and show how to achieve critical security and usability properties (safety, liveness, and paralysis-freeness) in a DASS. To illustrate the wide range of applications, we present three use cases of DASSes for improving digital asset custody: a multi-signature scheme that can "downgrade" the threshold should players become unavailable; a hybrid scheme where the centralized custodian can't refuse service; and a smart-contract-based scheme that supports recovery from unexpected bugs.