English

Graph Reconstruction from Noisy Random Subgraphs

Information Theory 2025-08-01 v2 Data Structures and Algorithms math.IT

Abstract

We consider the problem of reconstructing an undirected graph GG on nn vertices given multiple random noisy subgraphs or "traces". Specifically, a trace is generated by sampling each vertex with probability pvp_v, then taking the resulting induced subgraph on the sampled vertices, and then adding noise in the form of either (a) deleting each edge in the subgraph with probability 1pe1-p_e, or (b) deleting each edge with probability fef_e and transforming a non-edge into an edge with probability fef_e. We show that, under mild assumptions on pvp_v, pep_e and fef_e, if GG is selected uniformly at random, then O(pe1pv2logn)O(p_e^{-1} p_v^{-2} \log n) or O((fe1/2)2pv2logn)O((f_e-1/2)^{-2} p_v^{-2} \log n) traces suffice to reconstruct GG with high probability. In contrast, if GG is arbitrary, then exp(Ω(n))\exp(\Omega(n)) traces are necessary even when pv=1,pe=1/2p_v=1, p_e=1/2.

Keywords

Cite

@article{arxiv.2405.04261,
  title  = {Graph Reconstruction from Noisy Random Subgraphs},
  author = {Andrew McGregor and Rik Sengupta},
  journal= {arXiv preprint arXiv:2405.04261},
  year   = {2025}
}

Comments

6 pages, to appear in ISIT 2024

R2 v1 2026-06-28T16:19:23.667Z