Broadcasting in random recursive dags

📅 2023-06-02
🏛️ arXiv.org
📈 Citations: 2
Influential: 0
📄 PDF
🤖 AI Summary
This work investigates the solvability of root majority bit reconstruction in a random recursive $k$-DAG—constructed from $k$ initial root nodes, where each new node selects $k$ parents uniformly at random and inherits a bit via noisy majority voting (each parent’s bit is flipped independently with probability $p$). We establish the first exact phase transition characterization of noise tolerance for such $k$-DAGs: an explicit critical threshold $p_c(k) = frac{1}{2} - Theta(k^{-1/2})$ is derived. When $p < p_c(k)$, a global majority estimator achieves sub-majority error rate $c + o(1)$ for some $c < 1/2$; when $p > p_c(k)$, the error rate degrades to $1/2 + o(1)$, matching random guessing. This precisely identifies the fundamental information-theoretic limit for reliable broadcast reconstruction on this class of directed acyclic graphs.
📝 Abstract
A uniform $k$-{sc dag} generalizes the uniform random recursive tree by picking $k$ parents uniformly at random from the existing nodes. It starts with $k$ ''roots''. Each of the $k$ roots is assigned a bit. These bits are propagated by a noisy channel. The parents' bits are flipped with probability $p$, and a majority vote is taken. When all nodes have received their bits, the $k$-{sc dag} is shown without identifying the roots. The goal is to estimate the majority bit among the roots. We identify the threshold for $p$ as a function of $k$ below which the majority rule among all nodes yields an error $c+o(1)$ with $c<1/2$. Above the threshold the majority rule errs with probability $1/2+o(1)$.
Problem

Research questions and friction points this paper is trying to address.

Estimate majority bit in roots
Determine threshold for error probability
Analyze noisy bit propagation in dags
Innovation

Methods, ideas, or system contributions that make the work stand out.

Random recursive dags
Noisy channel propagation
Majority bit estimation
🔎 Similar Papers
No similar papers found.
S
Simon Briend
Université Paris-Saclay, CNRS, Laboratoire de Mathématiques d'Orsay, 91405, Orsay, France
L
L. Devroye
School of Computer Science, McGill University, Montreal, Canada
G
G. Lugosi
Department of Economics and Business, Pompeu Fabra University, Barcelona, Spain; ICREA, Pg. Lluís Companys 23, 08010 Barcelona, Spain; Barcelona School of Economics