TY - JOUR
T1 - Minimum restricted diameter spanning trees
AU - Hassin, Refael
AU - Levin, Asaf
PY - 2004/3/15
Y1 - 2004/3/15
N2 - Let G = (V,E) be a requirement graph. Let d = (dij)i,j = 1n be a length metric. For a tree T denote by d T(i,j) the distance between i and j in T (the length according to d of the unique i-j path in T). The restricted diameter of T, DT, is the maximum distance in T between pair of vertices with requirement between them. The minimum restricted diameter spanning tree problem is to find a spanning tree T such that the restricted diameter is minimized. We prove that the minimum restricted diameter spanning tree problem is NP-hard and that unless P=NP there is no polynomial time algorithm with performance guarantee of less than 2. In the case that G contains isolated vertices and the length matrix is defined by distances over a tree we prove that there exists a tree over the non-isolated vertices such that its restricted diameter is at most 4 times the minimum restricted diameter and that this constant is at least 312. We use this last result to present an O(log(n))-approximation algorithm.
AB - Let G = (V,E) be a requirement graph. Let d = (dij)i,j = 1n be a length metric. For a tree T denote by d T(i,j) the distance between i and j in T (the length according to d of the unique i-j path in T). The restricted diameter of T, DT, is the maximum distance in T between pair of vertices with requirement between them. The minimum restricted diameter spanning tree problem is to find a spanning tree T such that the restricted diameter is minimized. We prove that the minimum restricted diameter spanning tree problem is NP-hard and that unless P=NP there is no polynomial time algorithm with performance guarantee of less than 2. In the case that G contains isolated vertices and the length matrix is defined by distances over a tree we prove that there exists a tree over the non-isolated vertices such that its restricted diameter is at most 4 times the minimum restricted diameter and that this constant is at least 312. We use this last result to present an O(log(n))-approximation algorithm.
KW - Approximation algorithms
KW - Combinatorial optimization
KW - Minimum diameter spanning trees
UR - https://www.scopus.com/pages/publications/1242287857
U2 - 10.1016/S0166-218X(03)00360-3
DO - 10.1016/S0166-218X(03)00360-3
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
AN - SCOPUS:1242287857
SN - 0166-218X
VL - 137
SP - 343
EP - 357
JO - Discrete Applied Mathematics
JF - Discrete Applied Mathematics
IS - 3
ER -