Maintaining an EDCS in General Graphs: Simpler, Density-Sensitive and with Worst-Case Time Bounds

TL;DR

This paper simplifies and extends Bernstein and Stein's EDCS-based dynamic matching algorithm from bipartite to general graphs, achieving worst-case update times of O_epsilon(m^{1/4}) that are density-sensitive based on graph arboricity.

Contribution

The authors simplify the EDCS approach for bipartite graphs and generalize it to all graphs, maintaining worst-case update bounds and introducing density-sensitive performance.

Findings

Achieved worst-case update time of O_epsilon(m^{1/4}) for general graphs.

Extended the density-sensitive property based on graph arboricity.

Simplified the technical approach compared to prior work.

Abstract

In their breakthrough ICALP'15 paper, Bernstein and Stein presented an algorithm for maintaining a $(3/2+\epsilon)$-approximate maximum matching in fully dynamic {\em bipartite} graphs with a {\em worst-case} update time of $O_\epsilon(m^{1/4})$; we use the $O_\epsilon$ notation to suppress the $\epsilon$-dependence. Their main technical contribution was in presenting a new type of bounded-degree subgraph, which they named an {\em edge degree constrained subgraph (EDCS)}, which contains a large matching -- of size that is smaller than the maximum matching size of the entire graph by at most a factor of $3/2+\epsilon$. They demonstrate that the EDCS can be maintained with a worst-case update time of $O_\epsilon(m^{1/4})$, and their main result follows as a direct corollary. In their followup SODA'16 paper, Bernstein and Stein generalized their result for general graphs, achieving the same update time of $O_\epsilon(m^{1/4})$, albeit with an amortized rather than worst-case bound. To date, the best {\em deterministic} worst-case update time bound for {\em any} better-than-2 approximate matching is $O(\sqrt{m})$ [Neiman and Solomon, STOC'13], [Gupta and Peng, FOCS'13]; allowing randomization (against an oblivious adversary) one can achieve a much better (still polynomial) update time for approximation slightly below 2 [Behnezhad, Lacki and Mirrokni, SODA'20]. In this work we\footnote{\em quasi nanos, gigantium humeris insidentes} simplify the approach of Bernstein and Stein for bipartite graphs, which allows us to generalize it for general graphs while maintaining the same bound of $O_\epsilon(m^{1/4})$ on the {\em worst-case} update time. Moreover, our approach is {\em density-sensitive}: If the {\em arboricity} of the dynamic graph is bounded by $\alpha$ at all times, then the worst-case update time of the algorithm is $O_\epsilon(\sqrt{\alpha})$.