分布式算法课件4.ppt

Level k → level k+1 (cont’d) Each level-k component leader finds MWOE of its component. Combine level-k components using MWOEs, to obtain level-(k+1)components. Choose new leaders: For each new, level k+1 component, there is a unique edge e that is the MWOE of two level k sub-components Choose new leader to be the endpoint of e with the larger UID. Broadcast leader UID to new (merged) component. GHS terminates when there are no more outgoing edges. * Note on Synchronization This simplified version of GHS is designed to work with component levels synchronized. Difficulties can arise when they get out of synch. In particular, test messages are supposed to compare leader UIDs to determine whether endpoints are in the same component. Requires that the node being queried has up-to-date UID information. * Minimum spanning tree * a j i h g f e d c b k 0 8 5 12 1 2 3 4 9 6 7 10 13 11 Minimum spanning tree * a j i h g f e d c b k 0 8 5 12 1 2 3 4 9 6 7 10 13 11 Minimum spanning tree * a j i h g f e d c b k 0 8 5 12 1 2 3 4 9 6 7 10 13 11 Minimum spanning tree * a j i h g f e d c b k 0 8 5 12 1 2 3 4 9 6 7 10 13 11 Minimum spanning tree * a j i h g f e d c b k 0 8 5 12 1 2 3 4 9 6 7 10 13 11 Minimum spanning tree * a j i h g f e d c b k 0 8 5 12 1 2 3 4 9 6 7 10 13 11 Minimum spanning tree * a j i h g f e d c b k 0 8 5 12 1 2 3 4 9 6 7 10 13 11 9 11 Minimum spanning tree * a j i h g f e d c b k 0 8 5 12 1 2 3 4 9 6 7 10 13 11 Minimum spanning tree * a j i h g f e d c b k 0 8 5 12 1 2 3 4 9 6 7 10 13 11 Minimum spanning tree * a j i h g f e d c b k 0 8 5 12 1 2 3 4 9 6 7 10 13 11 Minimum spanning tree * a j i h g f e d c b k 0 8 5 12 1 2 3 4 9 6 7 10 13 11 Simplified GHS MST Algorithm Complexity: Time: O(n log n) n rounds for each level log n levels, because there are ≥ 2k nodes in each level k component. Messages: O( (n + |E|) log n) Na?ve analysis. At each level, O(n) messages sent on tree edges, O(|E|) messages overall for all the test messages and their re