English

Self-Healing Computation

Distributed, Parallel, and Cluster Computing 2014-10-28 v2 Cryptography and Security

Abstract

In the problem of reliable multiparty computation (RC), there are nn parties, each with an individual input, and the parties want to jointly compute a function ff over nn inputs. The problem is complicated by the fact that an omniscient adversary controls a hidden fraction of the parties. We describe a self-healing algorithm for this problem. In particular, for a fixed function ff, with nn parties and mm gates, we describe how to perform RC repeatedly as the inputs to ff change. Our algorithm maintains the following properties, even when an adversary controls up to t(14ϵ)nt \leq (\frac{1}{4} - \epsilon) n parties, for any constant ϵ>0\epsilon >0. First, our algorithm performs each reliable computation with the following amortized resource costs: O(m+nlogn)O(m + n \log n) messages, O(m+nlogn)O(m + n \log n) computational operations, and O()O(\ell) latency, where \ell is the depth of the circuit that computes ff. Second, the expected total number of corruptions is O(t(logm)2)O(t (\log^{*} m)^2), after which the adversarially controlled parties are effectively quarantined so that they cause no more corruptions.

Keywords

Cite

@article{arxiv.1405.1167,
  title  = {Self-Healing Computation},
  author = {George Saad and Jared Saia},
  journal= {arXiv preprint arXiv:1405.1167},
  year   = {2014}
}

Comments

17 pages and 1 figure. It is submitted to SSS'14

R2 v1 2026-06-22T04:06:54.683Z