LIPIcs.ESA.2022.8.pdf
- Filesize: 483 kB
- 4 pages
We revisit the classical edge coloring problem for general graphs in the streaming model. In this model, the input graph is presented as a stream of edges, and the algorithm must report colors assigned to the edges in a streaming fashion, using a memory of size O(n polylog n) on graphs of up to O(n²) edges. In ESA 2019 and SOSA 2021, two elegant randomized algorithms were presented for this problem in the adversarial edge arrival model, where the latest one colors any input graph using O(Δ²/s) colors with high probability in Õ(ns) space. In this short note, we propose two extremely simple streaming algorithms that achieve the same color and space bounds deterministically. Besides being surprisingly simple, our algorithms do not use randomness at all, and are very simple to analyze.
Feedback for Dagstuhl Publishing