A Deterministic Algorithm for the MST Problem in Constant Rounds of Congested Clique

TL;DR

This paper presents the first deterministic constant-round algorithm for solving the Minimum Spanning Tree problem in the Congested Clique model, improving over previous randomized solutions and applicable to related distributed models.

Contribution

It introduces a deterministic $ ext{O}(1)$ round algorithm for MST in the Congested Clique, resolving an open question and surpassing prior randomized approaches.

Findings

Deterministic $ ext{O}(1)$ rounds suffice for MST in the Congested Clique.

The algorithm's communication complexity makes it applicable to MPC variants.

It closes the gap between randomized and deterministic MST algorithms in this model.

Abstract

In this paper, we show that the Minimum Spanning Tree problem can be solved \emph{deterministically}, in $\mathcal{O}(1)$ rounds of the $\mathsf{Congested}$ $\mathsf{Clique}$ model. In the $\mathsf{Congested}$ $\mathsf{Clique}$ model, there are $n$ players that perform computation in synchronous rounds. Each round consist of a phase of local computation and a phase of communication, in which each pair of players is allowed to exchange $\mathcal{O}(\log n)$ bit messages. The studies of this model began with the MST problem: in the paper by Lotker et al.[SPAA'03, SICOMP'05] that defines the $\mathsf{Congested}$ $\mathsf{Clique}$ model the authors give a deterministic $\mathcal{O}(\log \log n)$ round algorithm that improved over a trivial $\mathcal{O}(\log n)$ round adaptation of Bor\r{u}vka's algorithm. There was a sequence of gradual improvements to this result: an $\mathcal{O}(\log \log \log n)$ round algorithm by Hegeman et al. [PODC'15], an $\mathcal{O}(\log^* n)$ round algorithm by Ghaffari and Parter, [PODC'16] and an $\mathcal{O}(1)$ round algorithm by Jurdzi\'nski and Nowicki, [SODA'18], but all those algorithms were randomized, which left the question about the existence of any deterministic $o(\log \log n)$ round algorithms for the Minimum Spanning Tree problem open. Our result resolves this question and establishes that $\mathcal{O}(1)$ rounds is enough to solve the MST problem in the $\mathsf{Congested}$ $\mathsf{Clique}$ model, even if we are not allowed to use any randomness. Furthermore, the amount of communication needed by the algorithm makes it applicable to some variants of the $\mathsf{MPC}$ model.