DocumentCode :
655169
Title :
Approximating Minimum-Cost k-Node Connected Subgraphs via Independence-Free Graphs
Author :
Cheriyan, Jini ; Vegh, Laszlo A.
Author_Institution :
Dept. of Combinatorics & Optimization, Univ. of Waterloo, Waterlooz, ON, Canada
fYear :
2013
fDate :
26-29 Oct. 2013
Firstpage :
30
Lastpage :
39
Abstract :
We present a 6-approximation algorithm for the minimum-cost k-node connected spanning sub graph problem, assuming that the number of nodes is at least k3(k-1)+k. We apply a combinatorial preprocessing, based on the Frank-Tardos algorithm for k-out connectivity, to transform any input into an instance such that the iterative rounding method gives a 2-approximation guarantee. This is the first constant-factor approximation algorithm even in the asymptotic setting of the problem, that is, the restriction to instances where the number of nodes is lower bounded by a function of k.
Keywords :
approximation theory; graph theory; iterative methods; 2-approximation algorithm; 6-approximation algorithm; Frank-Tardos algorithm; combinatorial preprocessing algorithm; constant-factor approximation algorithm; independence-free graphs; iterative rounding method; linear programming; minimum-cost k-node connected subgraph problem; Approximation algorithms; Graph connectivity; Iterative rounding; Linear Programming;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Foundations of Computer Science (FOCS), 2013 IEEE 54th Annual Symposium on
Conference_Location :
Berkeley, CA
ISSN :
0272-5428
Type :
conf
DOI :
10.1109/FOCS.2013.12
Filename :
6686138
Link To Document :
بازگشت