| 摘要 |
A least-cost path between a pair of nodes A and B of a network is realized by segmenting the network into three or more segments, with nodes A and B being in different ones of the segments, with the third or more other segments being interposed between the segments to which nodes A and B belong, concurrently process the different segment to identify paths and their associated costs, and processing results generated by the different processing units to compute the least-cost path. Advantageously, the segmenting is chosen to equalize the work load of the different processing units so that the units finish their task roughly at the same time.
|
