Randomized Algorithms for Determining the Majority on Graphs
Every node of an undirected connected graph is colored white or black. Adjacent nodes can be compared and the outcome of each comparison is either 0 (same color) or 1 (different colors). The aim is to discover a node of the majority color, or to conclude that there is the same number of black and white nodes. We consider randomized algorithms for this task and establish upper and lower bounds on their expected running time. Our main contribution are lower bounds showing that some simple and natural algorithms for this problem cannot be improved in general.
KeywordsConnected Graph Deterministic Algorithm Randomize Algorithm White Node Expander Graph
Unable to display preview. Download preview PDF.
- 1.Aigner, M.: Variants of the majority problem. Discrete Applied Mathematics (to appear) Google Scholar
- 7.Kutten, S., Peleg, D.: Fault-local distributed mending. In: Proc. 14th ACM Symposium on Principles of Distributed Computing, pp. 20–27 (1995)Google Scholar
- 8.Malek, M.: A comparison connection assignment for diagnosis of multiprocessor systems. In: Proc. 7th Symp. Comput. Architecture, pp. 31–35 (1980)Google Scholar
- 11.Yao, A.C.-C.: Probabilistic computations: Towards a unified measure of complexity. In: Proc. 18th Ann. IEEE Symp. on Foundations of Computer Science, pp. 222–227 (1977)Google Scholar