lec18-steiner

dynamic programming approach note that d6 compute a

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: e MST Weight ω(i,j) of each edge is a 4-tuple w(i,j)= (dij, -|yi-yj|, -max(yi,yj), -max(xi,xj)) Weights are compared under lexicographic ordering j (xj,yj) dij i (xi,yi) Use Prim’s algorithm to compute a MST based on the weight function we obtain a rectilinear MST since the 1st component of w(i,j) is dij This MST is separable since the next three components in w(i,j) help break ties Run time O(n2) Compute an Optimal L-shaped Mapping Choose an arbitrary node as root Using dynamic programming Compute at each node v the following Φl(vi): min cost of Ti with (v,vi) using lower L-shape Φu(vi): min cost of Ti with (v, vi) using upper L-shape Ti Ti Compute an Optimal L-shaped Mapping (Cont’d) Φl(v) (or Φu(v) ) can be computed by examining 2d combinations of Φl(vi) and Φu(vi) (1≤i≤d) with (pv,v) taking upper (or lower) L pv pv v T1 T2 v Td T1 T2 Td Applying Spanning/Steiner Tree Algorithms General cell design: channel intersection graphs Standard cell/Gate array/Sea-of-gate design rectilinear steiner/spanning trees or grid graphs...
View Full Document

This document was uploaded on 04/07/2014.

Ask a homework question - tutors are online