Sharding is an emerging technique to overcome scalability issues on blockchain based public ledgers. Without sharding, every node in the network has to listen to and process all ledger protocol messages. The basic idea of sharding is to parallelize the ledger protocol: the nodes are divided into smaller subsets that each take care of a fraction of the original load by executing lighter instances of the ledger protocol, also called shards. The smaller the shards, the higher the efficiency, as by increasing parallelism there is less overhead in the shard consensus.
In this vein, we propose a novel approach that leverages the sharding safety-liveness dichotomy. We separate the liveness and safety in shard consensus, allowing us to dynamically tune shard parameters to achieve essentially optimal efficiency for the current corruption ratio of the system. We start by sampling a relatively small shard (possibly with a small honesty ratio), and we carefully trade-off safety for liveness in the consensus mechanism to tolerate small honesty without losing safety. However, for a shard to be live, a higher honesty ratio is required in the worst case. To detect liveness failures, we use a so-called control chain that is always live and safe. Shards that are detected to be not live are resampled with increased shard size and liveness tolerance until they are live, ensuring that all shards are always safe and run with optimal efficiency. As a concrete example, considering a population of 10K parties, 30% corruption and 60-bit security, our design permits shards of size 200 parties in contrast to 6K parties in previous designs.
Moreover, in this highly concurrent execution setting, it is paramount to guarantee that both the sharded ledger protocol and its sub protocols (e.g., the shards) are secure under composition. To prove the security of our approach, we present ideal functionalities capturing a sharded ledger as well as ideal functionalities capturing the control chain and individual shard consensus, which needs adjustable liveness. We further formalize our protocols and prove that they securely realize the sharded ledger functionality in the UC framework.
Everlasting security models the setting where hardness assumptions hold during the execution of a protocol but may get broken in the future. Due to the strength of this adversarial model, achieving any meaningful security guarantees for composable protocols is impossible without relying on hardware assumptions (Müller-Quade and Unruh, JoC’10). For this reason, a rich line of research has tried to leverage physical assumptions to construct well-known everlasting cryptographic primitives, such as commitment schemes. The only known everlastingly UC secure commitment scheme, due to Müller-Quade and Unruh (JoC’10), assumes honestly generated hardware tokens. The authors leave the possibility of constructing everlastingly UC secure commitments from malicious hardware tokens as an open problem.
In this work we close this gap by presenting the first construction of an everlastingly UC-secure commitment scheme in the fully malicious token model. Our scheme assumes the existence of physically uncloneable functions (PUFs) and is secure in the common reference string model. We also show that our results are tight by giving an impossibility proof for everlasting UC-secure computation from non-erasable tokens (such as PUFs), even with trusted setup.
Craig Gentry and Shai Halevi and Bernardo Magri and Jesper Buus Nielsen and Sophia Yakoubov
Abstract: Private information retrieval (PIR) lets a client retrieve an entry from a database held by a server, without the server learning which entry was retrieved. Here we study a weaker variant that we call *random-index PIR* (RPIR). It differs from standard PIR in that the retrieved index is an output rather than an input of the protocol, and it is chosen at random.
Our motivation for studying RPIR comes from a recent work of Benhamouda et al. (TCC’20) about maintaining secret values on public blockchains. Their solution involves choosing a small anonymous committee from among a large universe, and here we show that RPIR can be used for that purpose.
The RPIR client must be implemented via secure MPC for this use case, stressing the need to make it as efficient as can be. Combined with recent techniques for secure-MPC with stateless parties, our results yield a new secrets-on-blockchain construction (and more generally large-scale MPC). Our solution tolerates any fraction f⪇1/2 of corrupted parties, solving an open problem left from the work of Benhamouda et al.
Considering RPIR as a primitive, we show that it is in fact equivalent to PIR when there are no restrictions on the number of communication rounds. On the other hand, RPIR can be implemented in a “noninteractive” setting, which is clearly impossible for PIR. We also study batch RPIR, where multiple indexes are retrieved at once. Specifically we consider a weaker security guarantee than full RPIR, which is still good enough for our motivating application. We show that this weaker variant can be realized more efficiently than standard PIR or RPIR, and we discuss one protocol in particular that may be attractive for practical implementations.