Distributed Graph Algorithms
This chapter addresses three basic graph problems encountered in the context of distributed systems. These problems are (a) the computation of the shortest paths between a pair of processes where a positive length (or weight) is attached to each communication channel, (b) the coloring of the vertices (processes) of a graph in Δ+1 colors (where Δ is the maximal number of neighbors of a process, i.e., the maximal degree of a vertex when using the graph terminology), and (c) the detection of knots and cycles in a graph. As for the previous chapter devoted to graph traversal algorithms, an aim of this chapter is not only to present specific distributed graph algorithms, but also to show that their design is not always obtained from a simple extension of their sequential counterparts.
KeywordsDistributed graph algorithm Cycle detection Graph coloring Knot detection Maximal independent set Problem reduction Shortest path computation
- 203.S. Kanchi, D. Vineyard, An optimal distributed algorithm for all-pairs shortest-path. Int. J. Inf. Theories Appl. 11(2), 141–146 (2004) Google Scholar
- 373.S. Toueg, An all-pairs shortest paths distributed algorithm. IBM Technical Report RC 8327, 1980 Google Scholar