Random Optimal Net Oriented PWB Design Automation.

Abstract

Presented is a new automated design approach for the net initial assignment, placement and routing of components on a printed wiring board (PWB). In order to obtain shortest total wire length, all nets under consideration are randomly connected as minimum Steiner trees in every process. The random optimal net process theory is used to eliminate redundant comparisons and the indirective optimal technique is wed to avoid calculating the exact length of minimum Steiner trees in the iterative process. Thus complex PWBs can be designed more efficiently. In the routing process, a modified Lee's algorithm is used to build the minimum spanning trees related to the minimum Steiner trees imagined in the placement process and search proper channels for the Steiner trees in terms of the minimum spanning trees.