Mittagsseminar (in cooperation with J. Lengler, A. Steger, and D. Steurer)

Date and Time: Thursday, April 29, 2021, 12:15 pm

Duration: 30 minutes

Location: Zoom: conference room

Speaker: Sebastian Brandt

Improved Distributed Lower Bounds for MIS and Bounded Degree Dominating Sets on Trees

Given a graph G = (V,E) and a parameter k, a k-degree dominating set is a subset S of V such that S is a dominating set of G and in the induced subgraph G[S] every node has degree at most k. For k = 0, this definition coincides with the definition of a maximal independent set (MIS), while for k > 0, the problem of finding a k-degree dominating set is a natural relaxation of the MIS problem. Due to work by Barenboim and Elkin [PODC'08] and Ghaffari [SODA'16], it is known that, on trees, an MIS, and therefore also a k-degree dominating set for any k, can be computed in O(log n / log log n) rounds deterministically, and in O(sqrt(log n)) rounds randomized, in the distributed LOCAL model.

In this talk, we present Omega(sqrt(log n))-round lower bounds for the deterministic, and Omega(sqrt(log log n))-round lower bounds for the randomized computation of k-degree dominating sets on trees, both in the LOCAL model, improving, for k > 0, on the best previously known lower bounds of Omega(log*n)) due to Linial [FOCS'87] and Naor [J.Disc.Math.'91]. Our results also slightly improve on the best previously known lower bounds for MIS on trees due to Balliu, Brandt, and Olivetti [FOCS'20], while providing a simpler proof.

This is joint work with Alkida Balliu, Fabian Kuhn, and Dennis Olivetti.

Upcoming talks     |     All previous talks     |     Talks by speaker     |     Upcoming talks in iCal format (beta version!)

Previous talks by year:   2024  2023  2022  2021  2020  2019  2018  2017  2016  2015  2014  2013  2012  2011  2010  2009  2008  2007  2006  2005  2004  2003  2002  2001  2000  1999  1998  1997  1996  

Information for students and suggested topics for student talks

Automatic MiSe System Software Version 1.4803M   |   admin login